Fraud prevention platform Sift raises $50M at over $1B valuation, eyes acquisitions

With the increase of digital transacting over the past year, cybercriminals have been having a field day.

In 2020, complaints of suspected internet crime surged by 61%, to 791,790, according to the FBI’s 2020 Internet Crime Report. Those crimes — ranging from personal and corporate data breaches to credit card fraud, phishing and identity theft — cost victims more than $4.2 billion.

For companies like Sift — which aims to predict and prevent fraud online even more quickly than cybercriminals adopt new tactics — that increase in crime also led to an increase in business.

Last year, the San Francisco-based company assessed risk on more than $250 billion in transactions, double from what it did in 2019. The company has over several hundred customers, including Twitter, Airbnb, Twilio, DoorDash, Wayfair and McDonald’s, as well a global data network of 70 billion events per month.

To meet the surge in demand, Sift said today it has raised $50 million in a funding round that values the company at over $1 billion. Insight Partners led the financing, which included participation from Union Square Ventures and Stripes.

While the company would not reveal hard revenue figures, President and CEO Marc Olesen said that business has tripled since he joined the company in June 2018. Sift was founded out of Y Combinator in 2011, and has raised a total of $157 million over its lifetime.

The company’s “Digital Trust & Safety” platform aims to help merchants not only fight all types of internet fraud and abuse, but to also “reduce friction” for legitimate customers. There’s a fine line apparently between looking out for a merchant and upsetting a customer who is legitimately trying to conduct a transaction.

Sift uses machine learning and artificial intelligence to automatically surmise whether an attempted transaction or interaction with a business online is authentic or potentially problematic.

Image Credits: Sift

One of the things the company has discovered is that fraudsters are often not working alone.

“Fraud vectors are no longer siloed. They are highly innovative and often working in concert,” Olesen said. “We’ve uncovered a number of fraud rings.”

Olesen shared a couple of examples of how the company thwarted fraud incidents last year. One recently involved money laundering through donation sites where fraudsters tested stolen debit and credit cards through fake donation sites at guest checkout.

“By making small donations to themselves, they laundered that money and at the same tested the validity of the stolen cards so they could use it on another site with significantly higher purchases,” he said. 

In another case, the company uncovered fraudsters using Telegram, a social media site, to make services available, such as food delivery, with stolen credentials.

The data that Sift has accumulated since its inception helps the company “act as the central nervous system for fraud teams.” Sift says that its models become more intelligent with every customer that it integrates.

Insight Partners Managing Director Jeff Lieberman, who is a Sift board member, said his firm initially invested in Sift in 2016 because even at that time, it was clear that online fraud was “rapidly growing.” It was growing not just in dollar amounts, he said, but in the number of methods cybercriminals used to steal from consumers and businesses.

Sift has a novel approach to fighting fraud that combines massive data sets with machine learning, and it has a track record of proving its value for hundreds of online businesses,” he wrote via email.

When Olesen and the Sift team started the recent process of fundraising, Insight actually approached them before they started talking to outside investors “because both the product and business fundamentals are so strong, and the growth opportunity is massive,” Lieberman added.

“With more businesses heavily investing in online channels, nearly every one of them needs a solution that can intelligently weed out fraud while ensuring a seamless experience for the 99% of transactions or actions that are legitimate,” he wrote. 

The company plans to use its new capital primarily to expand its product portfolio and to scale its product, engineering and sales teams.

Sift also recently tapped Eu-Gene Sung — who has worked in financial leadership roles at Integral Ad Science, BSE Global and McCann — to serve as its CFO.

As to whether or not that meant an IPO is in Sift’s future, Olesen said that Sung’s experience of taking companies through a growth phase such as what Sift is experiencing would be valuable. The company is also for the first time looking to potentially do some M&A.

“When we think about expanding our portfolio, it’s really a buy/build partner approach,” Olesen said.

#airbnb, #artificial-intelligence, #board-member, #credit-card, #credit-card-fraud, #crime, #crimes, #cybercrime, #doordash, #federal-bureau-of-investigation, #food-delivery, #fraud, #funding, #fundings-exits, #identity-theft, #insight-partners, #jeff-lieberman, #machine-learning, #mcdonalds, #online-fraud, #private-equity, #recent-funding, #san-francisco, #sift, #startup, #startups, #stripes, #tc, #twilio, #union-square-ventures, #wayfair, #y-combinator

0

To ensure inclusivity, the Biden administration must double down on AI development initiatives

The National Security Commission on Artificial Intelligence (NSCAI) issued a report last month delivering an uncomfortable public message: America is not prepared to defend or compete in the AI era. It leads to two key questions that demand our immediate response: Will the U.S. continue to be a global superpower if it falls behind in AI development and deployment? And what can we do to change this trajectory?

Left unchecked, seemingly neutral artificial intelligence (AI) tools can and will perpetuate inequalities and, in effect, automate discrimination. Tech-enabled harms have already surfaced in credit decisions, health care services, and advertising.

To prevent this recurrence and growth at scale, the Biden administration must clarify current laws pertaining to AI and machine learning models — both in terms of how we will evaluate use by private actors and how we will govern AI usage within our government systems.

The administration has put a strong foot forward, from key appointments in the tech space to issuing an Executive Order on the first day in office that established an Equitable Data Working Group. This has comforted skeptics concerned both about the U.S. commitment to AI development and to ensuring equity in the digital space.

But that will be fleeting unless the administration shows strong resolve in making AI funding a reality and establishing leaders and structures necessary to safeguard its development and use.

Need for clarity on priorities

There has been a seismic shift at the federal level in AI policy and in stated commitments to equality in tech. A number of high profile appointments by the Biden administration — from Dr. Alondra Nelson as Deputy of OSTP, to Tim Wu at the NEC, to (our former senior advisor) Kurt Campbell at the NSC — signal that significant attention will be paid to inclusive AI development by experts on the inside.

The NSCAI final report includes recommendations that could prove critical to enabling better foundations for inclusive AI development, such as creating new talent pipelines through a U.S. Digital Service Academy to train current and future employees.

The report also recommends establishing a new Technology Competitiveness Council led by the Vice President. This could prove essential in ensuring that the nation’s commitment to AI leadership remains a priority at the highest levels. It makes good sense to have the administration’s leadership on AI spearheaded by VP Harris in light of her strategic partnership with the President, her tech policy savvy and her focus on civil rights.

The U.S. needs to lead by example

We know AI is powerful in its ability to create efficiencies, such as plowing through thousands of resumes to identify potentially suitable candidates. But it can also scale discrimination, such as the Amazon hiring tool that prioritized male candidates or “digital redlining” of credit based on race.

The Biden administration should issue an Executive Order (EO) to agencies inviting ideation on ways AI can improve government operations. The EO should also mandate checks on AI used by the USG to ensure it’s not spreading discriminatory outcomes unintentionally.

For instance, there must be a routine schedule in place where AI systems are evaluated to ensure embedded, harmful biases are not resulting in recommendations that are discriminatory or inconsistent with our democratic, inclusive values — and reevaluated routinely given that AI is constantly iterating and learning new patterns.

Putting a responsible AI governance system in place is particularly critical in the U.S. Government, which is required to offer due process protection when denying certain benefits. For instance, when AI is used to determine allocation of Medicaid benefits, and such benefits are modified or denied based on an algorithm, the government must be able to explain that outcome, aptly termed technological due process.

If decisions are delegated to automated systems without explainability, guidelines and human oversight, we find ourselves in the untenable situation where this basic constitutional right is being denied.

Likewise, the administration has immense power to ensure that AI safeguards by key corporate players are in place through its procurement power. Federal contract spending was expected to exceed $600 billion in fiscal 2020, even before including pandemic economic stimulus funds. The USG could effectuate tremendous impact by issuing a checklist for federal procurement of AI systems — this would ensure the government’s process is both rigorous and universally applied, including relevant civil rights considerations.

Protection from discrimination stemming from AI systems

The government holds another powerful lever to protect us from AI harms: its investigative and prosecutorial authority. An Executive Order instructing agencies to clarify applicability of current laws and regulations (e.g., ADA, Fair Housing, Fair Lending, Civil Rights Act, etc.) when determinations are reliant on AI-powered systems could result in a global reckoning. Companies operating in the U.S. would have unquestionable motivation to check their AI systems for harms against protected classes.

Low-income individuals are disproportionately vulnerable to many of the negative effects of AI. This is especially apparent with regard to credit and loan creation, because they are less likely to have access to traditional financial products or the ability to obtain high scores based on traditional frameworks. This then becomes the data used to create AI systems that automate such decisions.

The Consumer Finance Protection Bureau (CFPB) can play a pivotal role in holding financial institutions accountable for discriminatory lending processes that result from reliance on discriminatory AI systems. The mandate of an EO would be a forcing function for statements on how AI-enabled systems will be evaluated, putting companies on notice and better protecting the public with clear expectations on AI use.

There is a clear path to liability when an individual acts in a discriminatory way and a due process violation when a public benefit is denied arbitrarily, without explanation. Theoretically, these liabilities and rights would transfer with ease when an AI system is involved, but a review of agency action and legal precedent (or rather, the lack thereof) indicates otherwise.

The administration is off to a good start, such as rolling back a proposed HUD rule that would have made legal challenges against discriminatory AI essentially unattainable. Next, federal agencies with investigative or prosecutorial authority should clarify which AI practices would fall under their review and current laws would be applicable — for instance, HUD for illegal housing discrimination; CFPB on AI used in credit lending; and the Department of Labor on AI used in determinations made in hiring, evaluations and terminations.

Such action would have the added benefit of establishing a useful precedent for plaintiff actions in complaints.

The Biden administration has taken encouraging first steps signaling its intent to ensure inclusive, less discriminatory AI. However, it must put its own house in order by directing that federal agencies require the development, acquisition and use of AI — internally and by those it does business with — is done in a manner that protects privacy, civil rights, civil liberties, and American values.

#artificial-intelligence, #biden-administration, #column, #consumer-financial-protection-bureau, #department-of-labor, #government, #machine-learning, #opinion, #policy, #united-states

0

Medchart raises $17M to help businesses more easily access patient-authorized health data

Electronic health records (EHR) have long held promise as a means of unlocking new superpowers for caregiving and patients in the medical industry, but while they’ve been a thing for a long time, actually accessing and using them hasn’t been as quick to become a reality. That’s where Medchart comes in, providing access to health information between businesses, complete with informed patient consent, for using said data at scale. The startup just raised $17 million across Series A and seed rounds, led by Crosslink Capital and Golden Ventures, and including funding from Stanford Law School, rapper Nas and others.

Medchart originally started out as more of a DTC play for healthcare data, providing access and portability to digital health information directly to patients. It sprung from the personal experience of co-founders James Bateman and Derrick Chow, who both faced personal challenges accessing and transferring health record information for relatives and loved ones during crucial healthcare crisis moments. Bateman, Medchart’s CEO, explained that their experience early on revealed that what was actually needed for the model to scale and work effectively was more of a B2B approach, with informed patient consent as the crucial component.

“We’re really focused on that patient consent and authorization component of letting you allow your data to be used and shared for various purposes,” Bateman said in an interview. “And then building that platform that lets you take that data and then put it to use for those businesses and services, that we’re classifying as ‘beyond care.’ Whether those are our core areas, which would be with your, your lawyer, or with an insurance provider, or clinical researcher — or beyond that, looking at a future vision of this really being a platform to power innovation, and all sorts of different apps and services that you could imagine that are typically outside that realm of direct care and treatment.”

Bateman explained that one of the main challenges in making patient health data actually work for these businesses that surround, but aren’t necessarily a core part of a care paradigm, is delivering data in a way that it’s actually useful to the receiving party. Traditionally, this has required a lot of painstaking manual work, like paralegals poring over paper documents to find information that isn’t necessarily consistently formatted or located.

“One of the things that we’ve been really focused on is understanding those business processes,” Bateman said. “That way, when we work with these businesses that are using this data — all permissioned by the patient — that we’re delivering what we call ‘the information,’ and not just the data. So what are the business decision points that you’re trying to make with this data?”

To accomplish this, Medchart makes use of AI and machine learning to create a deeper understanding of the data set in order to be able to intelligently answer the specific questions that data requesters have of the information. Therein lies their longterm value, since once that understanding is established, they can query the data much more easily to answer different questions depending on different business needs, without needing to re-parse the data every single time.

“Where we’re building these systems of intelligence on top of aggregate data, they are fully transferable to making decisions around policies for, for example, life insurance underwriting, or with pharmaceutical companies on real world evidence for their phase three, phase four clinical trials, and helping those teams to understand, you know, the the overall indicators and the preexisting conditions and what the outcomes are of the drugs under development or whatever they’re measuring in their study,” Bateman said.”

According to Ameet Shah, Partner at co-lead investor for the Series A Golden Ventures, this is the key ingredient in what Medchart is offering that makes the company’s offering so attractive in terms of long-term potential.

“What you want is you both depth and breadth, and you need predictability — you need to know that you’re actually getting like the full data set back,” Shah said in an interview. “There’s all these point solutions, depending on the type of clinic you’re looking at, and the type of record you’re accessing, and that’s not helpful to the requester. Right now, you’re putting the burden on them, and when we looked at it, we were just like ‘Oh, this is just a whole bunch of undifferentiated heavy lifting that the entire health tech ecosystem is trying to like solve for. So if [Medchart] can just commoditize that and drive the cost down as low as possible, you can unlock all these other new use cases that never could have been done before.”

One recent development that positions Medchart to facilitate even more novel use cases of patient data is the 21st Century Cures Act, which just went into effect on April 5, provides patients with immediate access, without charge, to all the health information in their electronic medical records. That sets up a huge potential opportunity in terms of portability, with informed consent, of patient data, and Bateman suggests it will greatly speed up innovation built upon the type of information access Medchart enables.

“I think there’s just going to be an absolute explosion in this space over the next two to three years,” Bateman said. “And at Medchart, we’ve already built all the infrastructure with connections to these large information systems. We’re already plugged in and providing the data and the value to the end users and the customers, and I think now you’re going to see this acceleration and adoption and growth in this area that we’re super well-positioned to be able to deliver on.”

#artificial-intelligence, #crosslink-capital, #digital-health, #golden-ventures, #health, #healthcare-data, #informed-consent, #insurance, #machine-learning, #nas, #pharmaceutical, #tc

0

Apple’s new iMac finally gets an actually good webcam

Apple introduced new iMacs at its event on Tuesday, outfitted with its M1 processor and redesigned inside and out from the ground up. The hardware is impressive, but one of the biggest improvements for everyone’s Zoom-heavy life might be the webcam. Apple said it’s the “best camera ever in a Mac,” which honestly wouldn’t take much, but its specs suggest it actually is a big upgrade.

The camera finally achieves 1080p video capabilities, and Apple has also equipped it with a larger sensor that should provide greatly-improved low light performance. The M1 chip has better image signal processing capabilities, and uses computational video powers to correct and improve the image on the fly, which has brought benefits to the image quality even on existing MacBook Air and MacBook Pro hardware with the same old, bad webcam equipment.

That should mean this iMac actually has really good image quality — or at least not image quality you need to be embarrassed about. The on-board machine learning processor in the M1, which Apple calls the Neural Engine, will be working in real-time to optimize lighting and do noise reduction, too.

On top of the camera, Apple touts new beam forming mics in a three-mic array that will optimize audio, focusing on your voice and eliminating background noise. All told, this should finally be a Mac that provides a videoconferencing experience that doesn’t feel like it’s stuck in the early 2000s.

#apple, #apple-inc, #apple-spring-hardware-event-2021, #computers, #computing, #imac, #macbook, #macbook-air, #machine-learning, #steve-jobs, #tc, #teleconferencing, #webcam

0

Announcing our TC Sessions: SaaS virtual event happening October 27

Software-as-a-Service (SaaS) is now the default business model for most B2B and B2C software startups. And while it’s been around for a while now, its momentum keeps accelerating and the ecosystem continues to expand as technologists and marketers are getting more sophisticated about how to build and sell SaaS products. For all of them, we’re pleased to announced TechCrunch Sessions: SaaS 2021, a one-day virtual event that will examine the state of SaaS to help startup founders, developers and investors understand the state of play and what’s next.

The single-day event will take place 100% virtually on October 27 and will feature actionable advice, Q&A with some of SaaS’s biggest names, and plenty of networking opportunities. $75 Early Bird Passes are now on sale. Book your passes today to save $100 before prices go up.

We’re not quite ready to disclose our agenda yet, but you can expect a mix of superstars from across the industry, ranging from some of the largest tech companies to up-and-coming startups that are pushing the limits of SaaS.

The plan is to look at a broad spectrum of what’s happening in with B2B startups and give you actionable insights into how to build and/or improve your own product. If you’re just getting started, we want you to come away with new ideas for how to start your company and if you’re already on your way, then our sessions on scaling both your technology and marketing organization will help you to get to that $100 million annual run rate faster.

In addition to other founders, you’ll also hear from enterprise leaders who decide what to buy — and the mistakes they see startups make when they try to sell to them.

But SaaS isn’t only about managing growth — though ideally, that’s a problem founders will face sooner or later. Some of the other specific topics we will look at are how to keep your services safe in an ever-growing threat environment, how to use open source to your advantage and how to smartly raise funding for your company.

We will also highlight how B2B and B2C companies can handle the glut of data they now produce and use it to build machine learning models in the process. We’ll talk about how SaaS startups can both do so themselves and help others in the process. There’s nary a startup that doesn’t want to use some form of AI these days, after all.

And because this is 2021, chances are we’ll also talk about building remote companies and the lessons SaaS startups can learn from the last year of working through the pandemic.

Don’t miss out. Book your $75 Early Bird pass today and save $100.

#business-models, #computing, #enterprise, #entrepreneurship, #machine-learning, #private-equity, #saas, #software, #software-as-a-service, #startup-company, #tc

0

Cape Privacy announces $20M Series A to help companies securely share data

Cape Privacy, the early stage startup that wants to make it easier for companies to share sensitive data in a secure and encrypted way, announced a $20 million Series A today.

Evolution Equity Partners led the round with participation from new investors Tiger Global Management, Ridgeline Partners and Downing Lane. Existing investors Boldstart Ventures, Version One Ventures, Haystack, Radical Ventures and a slew of individual investors also participated. The company has now raised approximately $25 million including a $5 million seed investment we covered last June..

Cape Privacy CEO Ché Wijesinghe says that the product has evolved quite a bit since we last spoke. “We have really focused our efforts on encrypted learning, which is really the core technology, which was fundamental to allowing the multi-party compute capabilities between two organizations or two departments to work and build machine learning models on encrypted data,” Wijesinghe told me.

Wijesinghe says that a key business case involves a retail company owned by a private equity firm sharing data with a large financial services company, which is using the data to feed its machine learning models. In this case, sharing customer data, it’s essential to do it in a secure way and that is what Cape Privacy claims is its primary value prop.

He said that while the data sharing piece is the main focus of the company, it has data governance and compliance components to be sure that entities sharing data are doing so in a way that complies with internal and external rules and regulations related to the type of data.

While the company is concentrating on financial services for now because Wijesinghe has been working with these companies for years, he sees uses cases far beyond a single vertical including pharmaceuticals, government, healthcare telco and manufacturing.

“Every single industry needs this and so we look at the value of what Cape’s encrypted learning can provide as really being something that can be as transformative and be as impactful as what SSL was for the adoption of the web browser,” he said.

Richard Seewald, founding and managing partner at lead investor Evolution Equity Partners likes that ability to expand the product’s markets. “The application in Financial Services is only the beginning. Cape has big plans in life sciences and government where machine learning will help make incredible advances in clinical trials and counter-terrorism for example. We anticipate wide adoption of Cape’s technology across many use cases and industries,” he said.

The company has recently expanded to 20 people and Wijesinghe, who is half Asian, takes DEI seriously. “We’ve been very, very deliberate about our DEI efforts, and I think one of the things that we pride ourselves in is that we do foster a culture of acceptance, that it’s not just about diversity in terms of color, race, gender, but we just hired our first non binary employee,” he said,

Part of making people feel comfortable and included involves training so that fellow employees have a deeper understanding of the cultural differences. The company certainly has diversity across geographies with employees in 10 different time zones.

The company is obviously remote with a spread like that, but once the pandemic is over, Wijesinghe sees bringing people together on occasion with New York City as the hub for the company where people from all over the world can fly in and get together.

#cape-privacy, #data-sharing, #encryption, #enterprise, #evolution-equity-partners, #funding, #machine-learning, #recent-funding, #security, #startups

0

Data scientists: Bring the narrative to the forefront

By 2025, 463 exabytes of data will be created each day, according to some estimates. (For perspective, one exabyte of storage could hold 50,000 years of DVD-quality video.) It’s now easier than ever to translate physical and digital actions into data, and businesses of all types have raced to amass as much data as possible in order to gain a competitive edge.

However, in our collective infatuation with data (and obtaining more of it), what’s often overlooked is the role that storytelling plays in extracting real value from data.

The reality is that data by itself is insufficient to really influence human behavior. Whether the goal is to improve a business’ bottom line or convince people to stay home amid a pandemic, it’s the narrative that compels action, rather than the numbers alone. As more data is collected and analyzed, communication and storytelling will become even more integral in the data science discipline because of their role in separating the signal from the noise.

Data alone doesn’t spur innovation — rather, it’s data-driven storytelling that helps uncover hidden trends, powers personalization, and streamlines processes.

Yet this can be an area where data scientists struggle. In Anaconda’s 2020 State of Data Science survey of more than 2,300 data scientists, nearly a quarter of respondents said that their data science or machine learning (ML) teams lacked communication skills. This may be one reason why roughly 40% of respondents said they were able to effectively demonstrate business impact “only sometimes” or “almost never.”

The best data practitioners must be as skilled in storytelling as they are in coding and deploying models — and yes, this extends beyond creating visualizations to accompany reports. Here are some recommendations for how data scientists can situate their results within larger contextual narratives.

Make the abstract more tangible

Ever-growing datasets help machine learning models better understand the scope of a problem space, but more data does not necessarily help with human comprehension. Even for the most left-brain of thinkers, it’s not in our nature to understand large abstract numbers or things like marginal improvements in accuracy. This is why it’s important to include points of reference in your storytelling that make data tangible.

For example, throughout the pandemic, we’ve been bombarded with countless statistics around case counts, death rates, positivity rates, and more. While all of this data is important, tools like interactive maps and conversations around reproduction numbers are more effective than massive data dumps in terms of providing context, conveying risk, and, consequently, helping change behaviors as needed. In working with numbers, data practitioners have a responsibility to provide the necessary structure so that the data can be understood by the intended audience.

#column, #computing, #data, #data-management, #data-visualization, #developer, #ec-column, #ec-consumer-applications, #ec-enterprise-applications, #enterprise, #machine-learning, #peter-wang, #startups, #storytelling

0

Tecton teams with founder of Feast open source machine learning feature store

Tecton, the company that pioneered the notion of the machine learning feature store, has teamed up with the founder of the open source feature store project called Feast. Today the company announced the release of version 0.10 of the open source tool.

The feature store is a concept that the Tecton founders came up with when they were engineers at Uber. Shortly thereafter an engineer named Willem Pienaar read the founder’s Uber blog posts on building a feature store and went to work building Feast as an open source version of the concept.

“The idea of Tecton [involved bringing] feature stores to the industry, so we build basically the best in class, enterprise feature store. […] Feast is something that Willem created, which I think was inspired by some of the early designs that we published at Uber. And he built Feast and it evolved as kind of like the standard for open source feature stores, and it’s now part of the Linux Foundation,” Tecton co-founder and CEO Mike Del Balso explained.

Tecton later hired Pienaar, who is today an engineer at the company where he leads their open source team. While the company did not originally start off with a plan to build an open source product, the two products are closely aligned, and it made sense to bring Pienaar on board.

“The products are very similar in a lot of ways. So I think there’s a similarity there that makes this somewhat symbiotic, and there is no explicit convergence necessary. The Tecton product is a superset of what Feast has. So it’s an enterprise version with a lot more advanced functionality, but at Feast we have a battle-tested feature store that’s open source,” Pienaar said.

As we wrote in a December 2020 story on the company’s $35 million Series B, it describes a feature store as “an end-to-end machine learning management system that includes the pipelines to transform the data into what are called feature values, then it stores and manages all of that feature data and finally it serves a consistent set of data.”

Del Balso says that from a business perspective, contributing to the open source feature store exposes his company to a different group of users, and the commercial and open source products can feed off one another as they build the two products.

“What we really like, and what we feel is very powerful here, is that we’re deeply in the Feast community and get to learn from all of the interesting use cases […] to improve the Tecton product. And similarly, we can use the feedback that we’re hearing from our enterprise customers to improve the open source project. That’s the kind of cross learning, and ideally that feedback loop involved there,” he said.

The plan is for Tecton to continue being a primary contributor with a team inside Tecton dedicated to working on Feast. Today, the company is releasing version 0.10 of the project.

#artificial-intelligence, #developer, #enterprise, #feature-stores, #linux-foundation, #machine-learning, #open-source, #tc

0

BigEye (formerly Toro) scores $17M Series A to automate data quality monitoring

As companies create machine learning models, the operations team needs to ensure the data used for the model is of sufficient quality, a process that can be time consuming. BigEye (formerly Toro), an early stage startup is helping by automating data quality.

Today the company announced a $17 million Series A led Sequoia Capital with participation from existing investor Costanoa Ventures. That brings the total raised to $21 million with the $4 million seed, the startup raised last May.

When we spoke to BigEye CEO and co-founder Kyle Kirwan last May, he said the seed round was going to be focussed on hiring a team — they are 11 now — and building more automation into the product, and he says they have achieved that goal.

“The product can now automatically tell users what data quality metrics they should collect from their data, so they can point us at a table in Snowflake or Amazon Redshift or whatever and we can analyze that table and recommend the metrics that they should collect from it to monitor the data quality — and we also automated the alerting,” Kirwan explained.

He says that the company is focusing on data operations issues when it comes to inputs to the model such as the table isn’t updating when it’s supposed to, it’s missing rows or there are duplicate entries. They can automate alerts to those kinds of issues and speed up the process of getting model data ready for training and production.

Bogomil Balkansky, the partner at Sequoia who is leading today’s investment sees the company attacking an important part of the machine learning pipeline. “Having spearheaded the data quality team at Uber, Kyle and Egor have a clear vision to provide always-on insight into the quality of data to all businesses,” Balkansky said in a statement.

As the founding team begins building the company, Kirwan says that building a diverse team is a key goal for them and something they are keenly aware of.

“It’s easy to hire a lot of other people that fit a certain mold, and we want to be really careful that we’re doing the extra work to [understand that just because] it’s easy to source people within our network, we need to push and make sure that we’re hiring a team that has different backgrounds and different viewpoints and different types of people on it because that’s how we’re going to build the strongest team,” he said.

BigEye offers on prem and SaaS solutions, and while it’s working with paying customers like Instacart, Crux Informatics, and Lambda School, the product won’t be generally available until later in the year.

#artificial-intelligence, #bigeye, #enterprise, #funding, #machine-learning, #recent-funding, #sequoia-capital, #startups, #tc, #toro

0

Amira Learning raises $11M to put its AI-powered literacy tutor in post-COVID classrooms

School closures due to the pandemic have interrupted the learning processes of millions of kids, and without individual attention from teachers, reading skills in particular are taking a hit. Amira Learning aims to address this with an app that reads along with students, intelligently correcting errors in real time. Promising pilots and research mean the company is poised to go big as education changes, and it has raised $11M to scale up with a new app and growing customer base.

In classrooms, a common exercise is to have students read aloud from a storybook or worksheet. The teacher listens carefully, stopping and correcting students on difficult words. This “guided reading” process is fundamental for both instruction and assessment: it not only helps the kids learn, but the teacher can break the class up into groups with similar reading levels so she can offer tailored lessons.

“Guided reading is needs-based, differentiated instruction and in COVID we couldn’t do it,” said Andrea Burkiett, Director of Elementary Curriculum and Instruction at the Savannah-Chatham County Public School System. Breakout sessions are technically possible, “but when you’re talking about a kindergarten student who doesn’t even know how to use a mouse or touchpad, COVID basically made small groups nonexistent.”

Amira replicates the guided reading process by analyzing the child’s speech as they read through a story and identifying things like mispronunciations, skipped words, and other common stumbles. It’s based on research going back 20 years that has tested whether learners using such an automated system actually see any gains (and they did, though generally in a lab setting).

In fact I was speaking to Burkiett out of skepticism — “AI” products are thick on the ground and while it does little harm if one recommends you a recipe you don’t like, it’s a serious matter if a kid’s education is impacted. I wanted to be sure this wasn’t a random app hawking old research to lend itself credibility, and after talking with Burkiett and CEO Mark Angel I feel it’s quite the opposite, and could actually be a valuable tool for educators. But it needed to convince educators first.

Not a replacement but a force multiplier

“You have to start by truly identifying the reason for wanting to employ a tech tool,” said Burkiett. “There are a lot of tech tools out there that are exciting, fun for kids, etc, but we could use all of them and not impact growth or learning at all because we didn’t stop and say, this tool helps me with this need.”

Amira was decided on as one that addresses the particular need in the K-5 range of steadily improving reading level through constant practice and feedback.

“When COVID hit, every tech tool came out of the woodwork and was made free and available,” Burkiett recalled. “With Amira you’re looking at a 1:1 tutor at their specific level. She’s not a replacement for a teacher — though it has been that way in COVID — but beyond COVID she could become a force multiplier,” said Burkiett.

You can see the old version of Amira in action below, though it’s been updated since:

Testing Amira with her own district’s students, Burkiett replicated the results that have been obtained in more controlled settings: as much as twice or three times as much progress in reading level based on standard assessment tools, some of which are built into the teacher-side Amira app.

Naturally it isn’t possible to simply attribute all this improvement to Amira — there are other variables in play. But it appears to help and doesn’t hinder, and the effect correlates with frequency of use. The exact mechanism isn’t as important as the fact that kids learn faster when they use the app versus when they don’t, and furthermore this allows teachers to better allocate resources and time. A kid who can’t use it as often because their family shares a single computer is at a disadvantage that has nothing to do with their aptitude — but this problem can be detected and accounted for by the teacher, unlike a simple “read at home” assignment.

“Outside COVID we would always have students struggling with reading, and we would have parents with the money and knowledge to support their student,” Burkiett explained. “But now we can take this tool and offer it to students regardless of mom and dad’s time, mom and dad’s ability to pay. We can now give that tutor session to every single student.”

“Radically sub-optimal conditions”

This is familiar territory for CEO Mark Angel, though the AI aspect, he admits, is new.

“A lot of the Amira team came from Renaissance Learning. bringing fairly conventional edtech software into elementary school classrooms at scale. The actual tech we used was very simple compared to Amira — the big challenge was trying to figure out how to make applications work with the teacher workflow, or make them friendly and resilient when 6 year olds are your users,” he told me.

“Not to make it trite, but what we’ve learned is really just listen to teachers — they’re the super-users,” Angel continued. “And to design for radically sub-optimal conditions, like background noise, kids playing with the microphone, the myriad things that happen in real life circumstances.”

Once they were confident in the ability of the app to reliably decode words, the system was given three fundamental tasks that fall under the broader umbrella of machine learning.

The first is telling the difference between a sentence being read correctly and incorrectly. This can be difficult due to the many normal differences between speakers. Singling out errors that matter, versus simply deviation from an imaginary norm (in speech recognition that is often American English as spoken by white people) lets readers go at their own pace and in their own voice, with only actual issues like saying a silent k noted by the app.

(On that note, considering the prevalence of English language learners with accents, I asked about the company’s performance and approach there. Angel said they and their research partners went to great lengths to make sure they had a representative dataset, and that the model only flags pronunciations that indicate a word was not read or understood correctly.)

The second is knowing what action to take to correct an error. In the case of a silent k, it matters whether this is a first grader who is still learning spelling or a fourth grader who is proficient. And is this the first time they’ve made that mistake, or the tenth? Do they need an explanation of why the word is this way, or several examples of similar words? “It’s about helping a student at a moment in time,” Angel said, both in the moment of reading that word, and in the context of their current state as a learner.

Screenshot of a reading assessment in the app Amira.

Third is a data-based triage system that warns students and parents if a kid may potentially have a language learning disorder like dyslexia. The patterns are there in how they read — and while a system like Amira can’t actually diagnose, it can flag kids who may be high risk to receive a more thorough screening. (A note on privacy: Angel assured me that all information is totally private and by default is considered to belong to the district. “You’d have to be insane to take advantage of it. We’d be out of business in a nanosecond.”)

The $10M in funding comes at what could be a hockey-stick moment for Amira’s adoption. (The round was led by Authentic Ventures II, LP, with participation from Vertical Ventures, Owl Ventures, and Rethink Education.)

“COVID was a gigantic spotlight on the problem that Amira was created to solve,” Angel said. “We’ve always struggled in this country to help our children become fluent readers. The data is quite scary — more than two thirds of our 4th graders aren’t proficient readers, and those two thirds aren’t equally distributed by income or race. It’s a decades long struggle.”

Having basically given the product away for a year, the company is now looking at how to convert those users into customers. It seems like, just like the rest of society, “going back to normal” doesn’t necessarily mean going back to 2019 entirely. The lessons of the pandemic era are sticking.

“They don’t have the intention to just go back to the old ways,” Angel explained. “They’re searching for a new synthesis — how to incorporate tech, but do it in a classroom with kids elbow to elbow and interacting with teachers. So we’re focused on making Amira the norm in a post-COVID classroom.”

Part of that is making sure the app works with language learners at more levels and grades, so the team is working to expand its capabilities upwards to include middle school students as well as elementary. Another is building out the management side so that success at the classroom and district levels can be more easily understood.

Cartoon illustration of an adventurous looking woman in front of a jungle and zeppelin.

Amira’s appearance got an update in the new app as well.

The company is also launching a new app aimed at parents rather than teachers. “A year ago 100 percent of our usage was in the classroom, then 3 weeks later 100 percent of our usage was at home. We had to learn a lot about how to adapt. Out of that learning we’re shipping Amira and the Story Craft that helps parents work with their children.”

Hundreds of districts are on board provisionally, but decisions are still being kicked down the road as they deal with outbreaks, frustrated parents, and every other chaotic aspect of getting back to “normal.”

Perhaps a bit of celebrity juice may help tip the balance in their favor. A new partnership with Houston Texans linebacker Brennan Scarlett has the NFL player advising the board and covering the cost of 100 students at a Portland, OR school through his education charity, the Big Yard Foundation — and more to come. It may be a drop in the bucket in the scheme of things, with a year of schooling disrupted, but teachers know that every drop counts.

#apps, #edtech, #education, #education-tech, #funding, #fundings-exits, #learning, #machine-learning, #reading, #recent-funding, #renaissance-learning, #schools, #startups, #tc

0

Pale Blue Dot aims to be Europe’s premier early-stage climate investor and has $100 million to prove it

When Hampus Jakobsson, Heidi Lindvall, and Joel Larsson, all well-known players in the European venture ecosystem, began talking about their new firm Pale Blue Dot, they began by looking at the problems with venture capital.

For the three entrepreneurs and investors, whose resumes included co-founding companies and accelerators like The Astonishing Tribe (Jakobsson) and Fast Track Malmö (Lindvall and Larsson) and working as a venture partner at BlueYard Capital (Jakobsson again), the problems were clear.

Their first thesis was that all investment funds should be impact funds, and be taking into account ways to effect positive change; their second thesis was that since all funds should be impact funds, what would be their point of differentiation — that is, where could they provide the most impact.

The three young investors hit on climate change as the core mission and ran with it.

As it was closing on €53 million ($63.3 million) last year, the firm also made its first investments in Phytoform, a London headquartered company creating new crops using computational biology and synbio; Patch, a San Francisco-based carbon-offsetting platform that finances both traditional and frontier “carbon sequestration” methods; and 20tree.ai, an Amsterdam-based startup, using machine learning and satellite data to understand trees to lower the risk of forest fires and power outages.

Now they’ve raised another €34 million and seven more investments on their path to doing between 30 and 35 deals.

These investments primarily focus on Europe and include Veat, a European vegetarian prepared meal company; Madefrom, a still-in-stealth company angling to make everyday products more sustainable; HackYourCloset, a clothing rental company leveraging fast fashion to avoid landfilling clothes; Hier, a fresh food delivery service; Cirplus, a marketplace for recycled plastics trading; and Overstory, which aims to prevent wildfires by giving utilities a view into vegetation around their assets. 

The team expects to be primarily focused on Europe, with a few opportunistic investments in the U.S., and intends to invest in companies that are looking to change systems rather than directly affect consumer behavior. For instance, a Pale Blue Dot investment likely wouldn’t include e-commerce filters for more sustainable shopping, but potentially could include investments in sustainable consumer products companies.

The size of the firm’s commitments will range up to €1 million and will look to commit to a lot of investments. That’s by design, said Jakobsson. “Climate is so many different fields that we didn’t want to do 50% of the fund in food or 50% of the fund in materials,” he said. Also, the founders know their skillsets, which are primarily helping early stage entrepreneurs scale and making the right connections to other investors that can add value.

“In every deal we’ve gotten in co-investors that add particular, amazing, value while we still try to be the shepherds and managers and sherpas,” Jakobsson said. “We’re the ones that are going to protect the founder from the hell-rain of investor opinions.”

Another point of differentiation for the firm are its limited partners. Jakobsson said they rejected capital from oil companies in favor of founders and investors from the tech community that could add value. These include Prima Materia, the investment vehicle for Spotify founder Daniel Ek; the founders of Supercell, Zendesk, TransferWise and DeliveryHero are also backing the firm. So too, is Albert Wenger, a managing partner at Union Square Ventures.

The goal, simply, is to be the best early stage climate fund in Europe.

“We want to be the European climate fund,” Lindvall said. “This is where we can make most of the difference.” 

#albert-wenger, #amsterdam, #blueyard-capital, #corporate-finance, #daniel-ek, #economy, #entrepreneurship, #europe, #finance, #food, #hampus-jakobsson, #heidi-lindvall, #investment, #joel-larsson, #london, #machine-learning, #managing-partner, #money, #oil, #pale-blue-dot, #partner, #private-equity, #san-francisco, #spotify, #supercell, #tc, #transferwise, #union-square-ventures, #united-states, #venture-capital, #zendesk

0

China’s Xpeng in the race to automate EVs with lidar

Elon Musk famously said any company relying on lidar is “doomed.” Tesla instead believes automated driving functions are built on visual recognition and is even working to remove the radar. China’s Xpeng begs to differ.

Founded in 2014, Xpeng is one of China’s most celebrated electric vehicle startups and went public when it was just six years old. Like Tesla, Xpeng sees automation as an integral part of its strategy; unlike the American giant, Xpeng uses a combination of radar, cameras, high-precision maps powered by Alibaba, localization systems developed in-house, and most recently, lidar to detect and predict road conditions.

“Lidar will provide the 3D drivable space and precise depth estimation to small moving obstacles even like kids and pets, and obviously, other pedestrians and the motorbikes which are a nightmare for anybody who’s working on driving,” Xinzhou Wu, who oversees Xpeng’s autonomous driving R&D center, said in an interview with TechCrunch.

“On top of that, we have the usual radar which gives you location and speed. Then you have the camera which has very rich, basic semantic information.”

Xpeng is adding lidar to its mass-produced EV model P5, which will begin delivering in the second half of this year. The car, a family sedan, will later be able to drive from point A to B based on a navigation route set by the driver on highways and certain urban roads in China that are covered by Alibaba’s maps. An older model without lidar already enables assisted driving on highways.

The system, called Navigation Guided Pilot, is benchmarked against Tesla’s Navigate On Autopilot, said Wu. It can, for example, automatically change lanes, enter or exit ramps, overtake other vehicles, and maneuver another car’s sudden cut-in, a common sight in China’s complex road conditions.

“The city is super hard compared to the highway but with lidar and precise perception capability, we will have essentially three layers of redundancy for sensing,” said Wu.

By definition, NGP is an advanced driver-assistance system (ADAS) as drivers still need to keep their hands on the wheel and take control at any time (Chinese laws don’t allow drivers to be hands-off on the road). The carmaker’s ambition is to remove the driver, that is, reach Level 4 autonomy two to four years from now, but real-life implementation will hinge on regulations, said Wu.

“But I’m not worried about that too much. I understand the Chinese government is actually the most flexible in terms of technology regulation.”

The lidar camp

Musk’s disdain for lidar stems from the high costs of the remote sensing method that uses lasers. In the early days, a lidar unit spinning on top of a robotaxi could cost as much as $100,000, said Wu.

“Right now, [the cost] is at least two orders low,” said Wu. After 13 years with Qualcomm in the U.S., Wu joined Xpeng in late 2018 to work on automating the company’s electric cars. He currently leads a core autonomous driving R&D team of 500 staff and said the force will double in headcount by the end of this year.

“Our next vehicle is targeting the economy class. I would say it’s mid-range in terms of price,” he said, referring to the firm’s new lidar-powered sedan.

The lidar sensors powering Xpeng come from Livox, a firm touting more affordable lidar and an affiliate of DJI, the Shenzhen-based drone giant. Xpeng’s headquarters is in the adjacent city of Guangzhou about 1.5 hours’ drive away.

Xpeng isn’t the only one embracing lidar. Nio, a Chinese rival to Xpeng targeting a more premium market, unveiled a lidar-powered car in January but the model won’t start production until 2022. Arcfox, a new EV brand of Chinese state-owned carmaker BAIC, recently said it would be launching an electric car equipped with Huawei’s lidar.

Musk recently hinted that Tesla may remove radar from production outright as it inches closer to pure vision based on camera and machine learning. The billionaire founder isn’t particularly a fan of Xpeng, which he alleged owned a copy of Tesla’s old source code.

In 2019, Tesla filed a lawsuit against Cao Guangzhi alleging that the former Tesla engineer stole trade secrets and brought them to Xpeng. XPeng has repeatedly denied any wrongdoing. Cao no longer works at Xpeng.

Supply challenges

While Livox claims to be an independent entity “incubated” by DJI, a source told TechCrunch previously that it is just a “team within DJI” positioned as a separate company. The intention to distance from DJI comes as no one’s surprise as the drone maker is on the U.S. government’s Entity List, which has cut key suppliers off from a multitude of Chinese tech firms including Huawei.

Other critical parts that Xpeng uses include NVIDIA’s Xavier system-on-the-chip computing platform and Bosch’s iBooster brake system. Globally, the ongoing semiconductor shortage is pushing auto executives to ponder over future scenarios where self-driving cars become even more dependent on chips.

Xpeng is well aware of supply chain risks. “Basically, safety is very important,” said Wu. “It’s more than the tension between countries around the world right now. Covid-19 is also creating a lot of issues for some of the suppliers, so having redundancy in the suppliers is some strategy we are looking very closely at.”

Taking on robotaxis

Xpeng could have easily tapped the flurry of autonomous driving solution providers in China, including Pony.ai and WeRide in its backyard Guangzhou. Instead, Xpeng becomes their competitor, working on automation in-house and pledges to outrival the artificial intelligence startups.

“The availability of massive computing for cars at affordable costs and the fast dropping price of lidar is making the two camps really the same,” Wu said of the dynamics between EV makers and robotaxi startups.

“[The robotaxi companies] have to work very hard to find a path to a mass-production vehicle. If they don’t do that, two years from now, they will find the technology is already available in mass production and their value become will become much less than today’s,” he added.

“We know how to mass-produce a technology up to the safety requirement and the quarantine required of the auto industry. This is a super high bar for anybody wanting to survive.”

Xpeng has no plans of going visual-only. Options of automotive technologies like lidar are becoming cheaper and more abundant, so “why do we have to bind our hands right now and say camera only?” Wu asked.

“We have a lot of respect for Elon and his company. We wish them all the best. But we will, as Xiaopeng [founder of Xpeng] said in one of his famous speeches, compete in China and hopefully in the rest of the world as well with different technologies.”

5G, coupled with cloud computing and cabin intelligence, will accelerate Xpeng’s path to achieve full automation, though Wu couldn’t share much detail on how 5G is used. When unmanned driving is viable, Xpeng will explore “a lot of exciting features” that go into a car when the driver’s hands are freed. Xpeng’s electric SUV is already available in Norway, and the company is looking to further expand globally.

#alibaba, #artificial-intelligence, #asia, #automation, #automotive, #baic, #bosch, #cars, #china, #cloud-computing, #driver, #electric-car, #elon-musk, #emerging-technologies, #engineer, #founder, #huawei, #lasers, #li-auto, #lidar, #livox, #machine-learning, #nio, #norway, #nvidia, #qualcomm, #robotaxi, #robotics, #self-driving-cars, #semiconductor, #shenzhen, #tc, #tesla, #transport, #transportation, #u-s-government, #united-states, #wu, #xavier, #xiaopeng, #xpeng

0

Deeplite raises $6M seed to deploy ML on edge with fewer compute resources

One of the issues with deploying a machine learning application is that it tends to be expensive and highly compute intensive.  Deeplite, a startup based in Montreal, wants to change that by providing a way to reduce the overall size of the model, allowing it to run on hardware with far fewer resources.

Today, the company announced a $6 million seed investment. Boston-based venture capital firm PJC led the round with help from Innospark Ventures, Differential Ventures and Smart Global Holdings. Somel Investments, BDC Capital and Desjardins Capital also participated.

Nick Romano, CEO and co-founder at Deeplite, says that the company aims to take complex deep neural networks that require a lot of compute power to run, tend to use up a lot of memory, and can consume batteries at a rapid pace, and help them run more efficiently with fewer resources.

“Our platform can be used to transform those models into a new form factor to be able to deploy it into constrained hardware at the edge,” Romano explained. Those devices could be as small as a cell phone, a drone or even a Raspberry Pi, meaning that developers could deploy AI in ways that just wouldn’t be possible in most cases right now.

The company has created a product called Neutrino that lets you specify how you want to deploy your model and how much you can compress it to reduce the overall size and the resources required to run it in production. The idea is to run a machine learning application on an extremely small footprint.

Davis Sawyer, chief product officer and co-founder, says that the company’s solution comes into play after the model has been built, trained and is ready for production. Users supply the model and the data set and then they can decide how to build a smaller model. That could involve reducing the accuracy a bit if there is a tolerance for that, but chiefly it involves selecting a level of compression — how much smaller you can make the model.

“Compression reduces the size of the model so that you can deploy it on a much cheaper processor. We’re talking in some cases going from 200 megabytes down to on 11 megabytes or from 50 megabytes to 100 kilobytes,” Davis explained.

Rob May, who is leading the investment for PJC, says that he was impressed with the team and the technology the startup is trying to build.

“Deploying AI, particularly deep learning, on resource-constrained devices, is a broad challenge in the industry with scarce AI talent and know-how available. Deeplite’s automated software solution will create significant economic benefit as Edge AI continues to grow as a major computing paradigm,” May said in a statement.

The idea for the company has roots in the TandemLaunch incubator in Montreal. It launched officially as a company in mid-2019 and today has 15 employees with plans to double that by the end of this year. As it builds the company, Romano says the founders are focused on building a diverse and inclusive organization.

“We’ve got a strategy that’s going to find us the right people, but do it in a way that is absolutely diverse and inclusive. That’s all part of the DNA of the organization,” he said.

When it’s possible to return to work, the plan is to have offices in Montreal and Toronto that act as hubs for employees, but there won’t be any requirement to come into the office.

“We’ve already discussed that the general approach is going to be that people can come and go as they please, and we don’t think we will need as large an office footprint as we may have had in the past. People will have the option to work remotely and virtually as they see fit,” Romano said.

#artificial-intelligence, #developer, #edge-computing, #funding, #machine-learning, #neural-networks, #recent-funding, #rob-may, #startups, #tc

0

How to choose and deploy industry-specific AI models

As artificial intelligence becomes more advanced, previously cutting-edge — but generic — AI models are becoming commonplace, such as Google Cloud’s Vision AI or Amazon Rekognition.

While effective in some use cases, these solutions do not suit industry-specific needs right out of the box. Organizations that seek the most accurate results from their AI projects will simply have to turn to industry-specific models.

Any team looking to expand its AI capabilities should first apply its data and use cases to a generic model and assess the results.

There are a few ways that companies can generate industry-specific results. One would be to adopt a hybrid approach — taking an open-source generic AI model and training it further to align with the business’ specific needs. Companies could also look to third-party vendors, such as IBM or C3, and access a complete solution right off the shelf. Or — if they really needed to — data science teams could build their own models in-house, from scratch.

Let’s dive into each of these approaches and how businesses can decide which one works for their distinct circumstances.

Generic models alone often don’t cut it

Generic AI models like Vision AI or Rekognition and open-source ones from TensorFlow or Scikit-learn often fail to produce sufficient results when it comes to niche use cases in industries like finance or the energy sector. Many businesses have unique needs, and models that don’t have the contextual data of a certain industry will not be able to provide relevant results.

Building on top of open-source models

At ThirdEye Data, we recently worked with a utility company to tag and detect defects in electric poles by using AI to analyze thousands of images. We started off using Google Vision API and found that it was unable to produce our desired results — with the precision and recall values of the AI models completely unusable. The models were unable to read the characters within the tags on the electric poles 90% of the time because it didn’t identify the nonstandard font and varying background colors used in the tags.

So, we took base computer vision models from TensorFlow and optimized them to the utility company’s precise needs. After two months of developing AI models to detect and decipher tags on the electric poles, and another two months of training these models, the results are displaying accuracy levels of over 90%. These will continue to improve over time with retraining iterations.

Any team looking to expand its AI capabilities should first apply its data and use cases to a generic model and assess the results. Open-source algorithms that companies can start off with can be found on AI and ML frameworks like TensorFlow, Scikit-learn or Microsoft Cognitive Toolkit. At ThirdEye Data, we used convolutional neural network (CNN) algorithms on TensorFlow.

Then, if the results are insufficient, the team can extend the algorithm by training it further on their own industry-specific data.

#artificial-intelligence, #big-tech, #column, #ec-ai, #ec-column, #ec-how-to, #machine-learning, #tc

0

Austin’s newest unicorn: The Zebra raises $150M after doubling revenue in 2020

The Zebra, an Austin-based company that operates an insurance comparison site, has raised $150 million in a Series D round that propels it into unicorn territory.

Both the round size and valuation are a substantial bump from the $38.5 million Series C that Austin-based The Zebra raised in February of 2020. (The company would not disclose its valuation at that time, saying now only that its new valuation of over $1 billion is a “nice step up.”)

The Zebra also would not disclose the name of the firm that led its Series D round, but sources familiar with the deal said it was London-based Hedosophia. Existing backers Weatherford Capital and Accel also participated in the funding event.

The round size also is bigger than all of The Zebra’s prior rounds combined, bringing the company’s total raised to $261.5 million since its 2012 inception. Previous backers also include Silverton Partners, Ballast Point Ventures, Daher Capital, Floodgate Fund, The Zebra CEO Keith Melnick, KDT and others. 

According to Melnick, the round was all primary, and included no debt or secondary.

The Zebra started out as a site for people looking for auto insurance via its real-time quote comparison tool. The company partners with the top 10 auto insurance carriers in the U.S. Over time, it’s also “naturally” evolved to offer homeowners insurance with the goal of eventually branching out into renters and life insurance. It recently launched a dedicated home and auto bundled product, although much of its recent growth still revolves around its core auto offering, according to Melnick.

Like many other financial services companies, The Zebra has benefited from the big consumer shift to digital services since the beginning of the COVID-19 pandemic.

And we know this because the company is one of the few that are refreshingly open about their financials. The Zebra doubled its net revenue in 2020 to $79 million compared to $37 million in 2019, according to Melnick, who is former president of travel metasearch engine Kayak. March marked the company’s highest-performing month ever, he said, with revenue totaling $12.5 million — putting the company on track to achieve an annual run rate of $150 million this year. For some context, that’s up from $8 million in September of 2020 and $6 million in May of 2020.

Also, its revenue per applicant has grown at a clip of 100% year over year, according to Melnick. And The Zebra has increased its headcount to over 325, compared to about 200 in early 2020.

“We’ve definitely improved our relationships with carriers and seen more carrier participation as they continue to embrace our model,” Melnick said. “And we’ve leaned more into brand marketing efforts.”

The Zebra CEO Keith Melnick. Image courtesy of The Zebra

The company was even profitable for a couple of months last year, somewhat “unintentionally,” according to Melnick.

“We’re not highly unprofitable or burning through money like crazy,” he told TechCrunch. “This new raise wasn’t to fund operations. It’s more about accelerating growth and some of our product plans. We’re pulling forward things that were planned for later in time. We still had a nice chunk of money sitting on our balance sheet.”

The company also plans to use its new capital to do more hiring and focus strongly on continuing to build The Zebra’s brand, according to Melnick. Some of the things the company is planning include a national advertising campaign and adding tools and information so it can serve as an “insurance advisor,” and not just a site that refers people to carriers. It’s also planning to create more “personalized experiences and results” via machine learning.

“We are accelerating our efforts to make The Zebra a household name,” Melnick said. “And we want a deeper connection with our users.” It also aims to be there for a consumer through their lifecycle — as they move from being renters to homeowners, for example.

And while an IPO is not out of the question, he emphasizes that it’s not the company’s main objective at this time.

“I definitely try not to get locked on to a particular exit strategy. I just want to make sure we continue to build the best company we can. And then, I think the exit will make itself apparent,” Melnick said. “I’m not blind and am very aware that public market valuations are strong right now and that may be the right decision for us, but for now, that’s not the ultimate goal for me.”

To the CEO, there’s still plenty of runway.

“This is a big milestone, but I do feel like for us that this is just the beginning,” he said. “We’ve just scratched the surface of it.”

Early investor Mark Cuban believes the company is at an inflection point.

” ‘Startup’ isn’t the right word anymore,” he said in a written statement. “The Zebra is a full fledged tech company that is taking on – and solving – some of the biggest challenges in the $638B insurance industry.”

Accel Partner John Locke said the firm has tripled down on its investment in The Zebra because of its confidence in not only what the company is doing but also its potential.

“In an increasingly noisy insurance landscape that includes insurtechs and traditional carriers, giving consumers the ability to compare everything in one place is is more and more valuable,” he told TechCrunch. “I think The Zebra has really seized the mantle of becoming the go-to site for people to compare insurance and then that’s showing up in the numbers, referral traffic and fundraise interest.”

#accel, #animals, #austin, #auto-insurance, #ballast-point-ventures, #connect, #finance, #financial-services, #floodgate-fund, #funding, #fundings-exits, #hedosophia, #insurance, #insurtech, #john-locke, #life-insurance, #machine-learning, #mark-cuban, #model, #recent-funding, #silverton-partners, #startups, #the-zebra, #united-kingdom, #united-states, #venture-capital, #zebra

0

How we dodged risks and raised millions for our open-source machine language startup

Open-source software gave birth to a slew of useful software in recent years. Many of the great technologies that we use today were born out of open-source development: Android, Firefox, VLC media player, MongoDB, Linux, Docker and Python, just to name a few, with many of these also developing into very successful for-profit companies.

While there are some dedicated open-source investors such as the Apache Software Foundation incubator and OSS Capital, the majority of open-source companies will raise from traditional venture capital firms.

Our team has raised from traditional venture capital firms like Speedinvest, open-source-specific firms like OSS, and even from more hybrid firms like OpenOcean, which was created by the founders and senior leadership teams at MariaDB and MySQL. These companies understandably have a significant but not exclusive open-source focus.

Our area of innovation is an open-source AutoML server that reduces model training complexity and brings machine learning to the source of the data. Ultimately, we feel democratizing machine learning has the potential to truly transform the modern business world. As such, we successfully raised $5 million in seed funding to help bring our vision to the current marketplace.

Here, we aim to provide insights and advice for open-source startups that hope to follow a similar path for securing funding, and also detail some of the important risks your team needs to consider when crafting a business model to attract investment.

Strategies for acquiring open-source seed funding

Obviously, venture capitalists find many open-source software initiatives to be worthy investments. However, they need to understand any inherent risks involved when successfully commercializing an innovative idea. Finding low-risk investments that lead to lucrative business opportunities remains an important goal for these firms.

In our experience, we found these risks fall into three major categories: market risk, execution risk, and founders’ risk. Explaining all three to potential investors in a concise manner helps dispel their fears. In the end, low-risk, high-reward scenarios obviously attract tangible interest from sources of venture capital.

Ultimately, investment companies want startups to generate enough revenue to reach a valuation exceeding $1 billion. While that number is likely to increase over time, it remains a good starting point for initial funding discussions with investors. Annual revenue of $100 million serves as a good benchmark for achieving that valuation level.

Market risks in open-source initiatives

Market risks for open-source organizations tend to be different when compared to traditional businesses seeking funding. Notably, investors in these traditional startups are taking a larger leap of faith.

#artificial-intelligence, #column, #ec-ai, #ec-column, #ec-how-to, #entrepreneurship, #machine-learning, #open-source-software, #private-equity, #startups, #tc, #venture-capital

0

Watch a monkey equipped with Elon Musk’s Neuralink device play Pong with its brain

Elon Musk’s Neuralink, one of his many companies and the only one currently focused on mind control (that we’re aware of), has released a new blog post and video detailing some of its recent updates — including using its hardware to make it possible for a monkey to play pong with only its brain.

In the video above, Neuralink demonstrates how it used its sensor hardware and brain implant to record a baseline of activity from this macaque (named ‘Pager’) as it played a game on-screen where it had to move a token to different squares using a joystick with its hand. Using that baseline data, Neuralink was able to use machine learning to anticipate where Pager was going to be moving the physical controller, and was eventually able to predict it accurately before the move was actually made. Researchers then removed the paddle entirely, and eventually did the same thing with Pong, ultimately ending up at a place where Pager no longer was even moving its hand on the air on the nonexistent paddle, and was instead controlling the in-game action entirely with its mind via the Link hardware and embedded neural threads.

The last we saw of Neuralink, Musk himself was demonstrating the Link tech live in August 2020, using pigs to show how it was able to read signals from the brain depending on different stimuli. This new demo with Pager more clearly outlines the direction that the tech is headed in terms of human applications, since, as the company shared on its blog, the same technology could be used to help patients with paralysis manipulate a cursor on a computer, for instance. That could be applied to other paradigms as well, including touch controls on an iPhone, and even typing using a virtual keyboard, according to the company.

Musk separately tweeted that in fact, he expects the initial version of Neuralink’s product to be able to allow someone with paralysis that prevents standard modes of phone interaction to use one faster than people using their thumbs for input. He also added that future iterations of the product would be able to enable communication between Neuralinks in different parts of a patient’s body, transmitting between an in-brain node and neural pathways in legs, for instance, making it possible for “paraplegics to walk again.”

These are obviously bold claims, but the company cites a lot of existing research that undergirds its existing demonstrations and near-term goals. Musk’s more ambitious claims, should, like all of his projections, definitely be taken with a healthy dose of skepticism. He did add that he hopes human trials will begin to get underway “hopefully later this year,” for instance – which is already two years later than he was initially anticipating those might start.

#biotech, #bmi, #elon, #elon-musk, #iphone, #machine-learning, #musk, #neuralink, #pager, #paralysis, #pong, #science, #science-and-technology, #tc, #technology, #transhumanism

0

Mercato raises $26M Series A to help smaller grocers compete online

The pandemic upended the way people shop for their everyday needs, including groceries. Online grocery sales in the U.S. are expected to reach 21.5% of the total grocery sales by 2025, after leaping from 3.4% pre-pandemic to 10.2% as of 2020. One business riding this wave is Mercato, an online grocery platform that helps smaller grocers and speciality food stores get online quickly. After helping grow its merchant sales by 1,300% in 2020, Mercato has now closed on $26 million in Series A funding, the company tells TechCrunch.

The round was led by Velvet Sea Ventures with participation from Team Europe, the investing arm of Lukasz Gadowski, co-founder of Delivery Hero. Seed investors Greycroft and Loeb.nyc also returned for the new round Gadowski and Mike Lazerow of Velvet Sea Ventures have also now joined Mercato’s board.

Mercato itself was founded in 2015 by Bobby Brannigan, who had grown up helping at his family’s grocery store in Brooklyn. But instead of taking over the business, as his Dad had hoped, Brannigan left for college and eventually went on to bootstrap a college textbook marketplace, Valore Books, to $100 million in sales. After selling the business, he returned his focus to the family’s store and found that everything was still operating the way it had been decades ago.

Image Credits: Bobby Brannigan of Mercato

“He had a very basic website, no e-commerce, no social media, and no point-of-sale system,” explains Brannigan. “I said, ‘I’m going to build what you need.’ This was my opportunity to help my dad in an area that I knew about,” he adds.

Brannigan recruited some engineers from his last company to help him build the software systems to modernize his dad’s store, including Mercato’s co-founders Dave Bateman, Michael Mason, and Matthew Alarie. But the team soon realized could do more than help just Brannigan’s dad — they could also help the 40,000 independent grocery stores just like him better compete with the Amazon’s of the world.

The result was Mercato, a platform-as-a-service that makes it easier for smaller grocers and speciality food shops to go online to offer their inventory for pickup or delivery, without having to partner with a grocery delivery service like Instacart, AmazonFresh or Shipt.

The solution today includes an e-commerce website and data analytics platform that helps stores understand what their customers are looking for, where customers are located, how to price their products, and other insights that help them to better run their store. And Mercato is now working on adding on a supply platform to help the stores buy inventory through their system, Brannigan notes.

“Basically, the vision of it is to give them the tech, the systems, and the platform they need to be successful in this day and age,” notes Brannigan.

He likens Mercato as a sort of “Shopify for groceries,” as it gives stores their own page on Mercato where they can reach customers. When the customer visits Mercato on the web or via its app, they can enter in their zip code to see which local stores offer online shopping. Some stores simply redirect their existing websites to their Mercato page, as they can continue to offer other basic information, like address, hours, and other details about their stores on the Mercato-provided site, while gaining access to Mercato’s over 1 million customers.

However, merchants can also opt for a white-label solution that they can plug into their own website, which uses their own branding.

The stores can further customize the experience they want to offer customers, in terms of pickup and delivery, and the time frames for both they want to commit to. If they want to ease into online grocery, for example, they can start with next-day delivery services, then speed thing up to same-day when they’re ready. They can also set limits on how many time slots they offer per hour, based on staffing levels.

Image Credits: Mercato

Unlike Instacart and others which send shoppers to stores to fill the orders, Mercato allows the merchants themselves to maintain the customer relationship by handling the orders themselves, which they can receive via email, text or even robo-phone calls.

“They’re maintaining that relationship,” says Brannigan. “Usually, it’s a lot better if it’s somebody from the store [doing the shopping] because they might know the customer; they know the kind of product they’re looking for. And if they don’t have it, they know something else they can recommend — so they’re like a really efficient recommendation engine.”

“The big difference between an Instacart shopper and the worker in the store is that the worker in the store understands that somebody is trying to put a meal on the table, and certain items could be an important ingredient,” he notes. “For the shoppers at Instacart, it’s about a time clock: how quickly can they pick an order to make the most money.”

The company contracts with both national and regional couriers to handle the delivery portion, once orders are ready.

Mercato’s system was put to test during the pandemic, when demand for online grocery skyrocketed.

This is where Mercato’s ability to rapidly onboard merchants came in handy. The company says it can take stores online in just 24 hours, as it has built out a centralized product catalog of over a million items. It then connects with the store’s point-of-sale system, and uploads and matches the store’s products to their own database. This allows Mercato to map around 95% of the store’s products in a matter of minutes, with the last bit being added manually — which helps to build out Mercato’s catalog even further. Today, Mercato can integrate with virtually all point-of-sale (POS) solutions in the grocery market, which is more than 30 different systems.

As customers shop, Mercato’s system uses machine learning to help determine if a product is likely in stock by examining movement data.

“One of the challenges in grocery is that most stores actually don’t know how many quantities they have in stock of a product,” explains Brannigan. “So we launch a store, we integrate with the POS. And with the POS we can see how quickly a product is moving in-store and online. Based on movement, we can calculate what is in stock.”

This system, he says, continues to get smarter over time, too.

“We’re certainly three to five years ahead, and we’re not going back,” says Brannigan of the COVID impacts to the online grocery business. “It’s very plentiful now in many places, in terms of e-commerce offerings. And the nature of retail businesses is competitive. So if 1% of people are online, it might not drive other people. But if you have 15% of stores online, then other stores have to get online or they won’t be able to compete,” he notes.

Mercato generates revenue both from its consumer-facing membership program, with plans that range from $96/year – $228/year, depending on distance, and from the merchants themselves, who pay a single digit percentage transaction fee on orders — a lower percentage than what restaurant delivery companies charge.

The company has now scaled its service to over 1,000 merchants across 45 U.S. states, including big cities like New York, Chicago, L.A. D.C., Boston, Philadelphia, and others.

With the additional funding, Mercato aims to expand its remotely distributed team of now 80 employees, as well as its data analytics platform, which will help merchants make better decisions that impact their business. It also plans to refresh the consumer subscription to add more benefits and perks that make it more compelling.

Mercato declined to share its valuation or revenue, but as of the start of the pandemic last year, the company had said it was reaching a billion in sales and a $700 million run rate.

#e-commerce, #ecommerce, #funding, #grocery-store, #lukasz-gadowski, #machine-learning, #online-grocery, #online-shopping, #retailers, #shopping, #startups, #supermarkets, #velvet-sea-ventures

0

Comet announces $13M Series A for ML model building tool

There’s been a lot of investment in machine learning startups recently as companies try to push the notion into a wider variety of endeavors. Comet, a company that helps customers iterate on models in an experimentation process designed to eventually reach production, announced a $13 million Series A today.

Scale Venture Partners led the round with help from existing investors Trilogy Equity Partners and Two Sigma Ventures. The startup has raised almost $20 million, according to Crunchbase data.

Investors saw a company that has grown revenue over 500% over the last year, says Gideon Mendels, co-founder and CEO. “Things have been working very well for us. On the product side, we’ve continued to double down on what we call experimentation management where we are really tracking these models — data that came into them, the hyper parameters and helping teams to debug and understand what’s going on with their models,” he said.

In addition to the funding, the company is also announcing an expansion of the platform to follow the models into post production with a product they are calling Comet Model Production Monitoring (MPM).

“The model production monitoring product essentially focuses on models post production. The original product was more around how multiple offline experiments are modeled during training, while MPM is focused on these models once they hit production for the first time,” Mendels explained.

Andy Vitus, partner at investor Scale Venture Partners, sees model lifecycle management tooling like Comet’s as a developing market. “Machine learning and AI will drive the future of enterprise software, and ensuring that organizations have full visibility and control of a model’s life cycle will be imperative to it,” Vitus said in a statement

As the company grows, it’s opening a new engineering hub in Israel in addition to its office in NYC. While these offices are closed for now, Mendels says that they will have a hybrid office when the pandemic ebbs.

“Moving forward we are planning to have an office in New York City and another office in Tel Aviv. But we’re not going to require anyone to work from the office if they choose not to, or, they can come in a couple days a week. And we’re still going to support hires from around the world.”

#artificial-intelligence, #comet, #data-science, #funding, #machine-learning, #recent-funding, #scale-venture-partners, #startups, #tc

0

YC-backed Abacum nets $7M to empower finance teams with real-time data and collaboration tools

SaaS to support mid-sized companies’ financial planning with real-time data and native collaboration isn’t the sexiest startup pitch under the sun but it’s one that’s swiftly netted Abacum a bunch of notable backers — including Creandum, which is leading a $7M seed round that’s being announced today.

The rosters of existing investors also participating in the round are Y Combinator (Abacum was part of its latest batch), PROFounders, and K-Fund, along with angel investors such as Justin Kan (Atrium and Twitch co-founder and CEO); Maximilian Tayenthal (N26 co-founder and co-CEO & CFO); Thomas Lehrman (GLG co-founder and ex-CEO), Avi Meir (TravelPerk co-founder and CEO); plus Jenny Bloom (Zapier CFO and Mailchimp ex-CFO) and Mike Asher (CFO at Neo4j).

Abacum was founded last year in the middle of the COVID-19 global lockdown, after what it says was around a year of “deep research” to feed its product development. They launched their SaaS in June 2020. And while they’re not disclosing customer numbers at this early stage their first clients include a range of scale-up companies in the US and in Europe, including the likes of Typeform, Cabify, Ebury, Garten, Jeff and Talkable.

The startup’s Spanish co-founders — Julio Martinez, a fintech entrepreneur with an investment banking background, and Jorge Lluch, a European Space Agency engineer turned CFO/COO — spotted an opportunity to build dedicated software for mid-market finance teams to provide real-time access to data via native collaborative that plugs into key software platforms used by other business units, having felt the pain of a lack of access to real-time data and barriers to collaboration in their own professional experience with the finance function.

The idea with Abacum is to replace the need for finance teams to manually update their models. The SaaS automatically does the updates, fed with real-time data through direct integrations with software used by teams dealing with functions like HR, CRM, ERP (and so on) — empowering the finance function to collaborate more easily across the business and bolster its strategic decision-making capabilities.

The startup’s sales pitch to the target mid-sized companies is multi-layered. Abacum says its SaaS both saves finance teams time and enables faster-decision making.

“Prior to using Abacum, finance analysts in our clients were easily spending 50% to 70% of their time in manual tasks like downloading files from different systems, copy&pasting them in massive spreadsheets (that crash frequently), formatting the data by manually adding and removing rows, columns and formats, connecting the data in a model prone to manual error (e.g. vlookups & sumifs),” Martinez tells TechCrunch. “With Abacum, this entire manual part is automatically done and the finance professionals can spend their time analyzing and adding real value to the business.”

“We enable faster decisions that were not possible prior to Abacum. For instance, some of our clients were updating their cohort analysis on a quarterly basis only because the associated manual tasks were too painful. With us, they’re able to update the analysis weekly and take better decisions as a result.”

The SaaS also supports decisions in another way — by applying machine learning to business data to generate estimates on future performance, providing an AI-based reference point based on historical data that finance teams can use to inform their assumptions.

And it aids cross-business collaboration — allowing users to share and gather information “easily through workflows and permissions”. “We see that this results in faster and richer decisions as more stakeholders are brought into the process,” he adds.

Martinez says Abacum chose to focus on mid-market finance teams because they face “more challenges and inefficiencies” vs the smaller (and larger) ends of the market. “In that segment, the finance function is underinvested — they face the acute complexities of scaling companies that become very pressing but at the same time they are still considered a support function, a back-office,” he argues.

“Abacum makes finance a strategic function — we deliver native collaboration to finance teams so that they become the trusted business partner they want to be. We also see that the pandemic has accelerated the need for finance teams to collaborate effectively and work remotely,” he adds.

He also describes the mid market segment as “fairly unpenetrated” — claiming many companies do not yet having a solution in place.

While competitors he points to when asked about other players in the space are long in the tooth in digital terms: Adaptive Insights (2003); Host Analytics (2001); and Anaplan (2008).

Commenting on the seed round in a statement, Peter Specht, principal at Creandum, added: “The financial planning processes in many companies are ripe for disruption and demand more automation. Abacum’s slick solution empowers finance teams to be more collaborative, efficient and better informed with access to real-time data. We were impressed by their user-friendly product, the initial hiring of top talent, and crucially the strong founders and their extensive operational experience — including as CFOs and entrepreneurs who have experienced the problem first-hand. We are delighted to be part of Abacum’s journey to empower global SMEs to bring their financial operations to new levels.”

Abacum’s seed financing will be ploughed into product development and growth, per Martinez, who says it’s focused on wooing finance teams in the US and Europe for now.

#abacum, #adaptive-insights, #anaplan, #artificial-intelligence, #avi-meir, #business-models, #cfo, #chief-financial-officer, #creandum, #crm, #europe, #finance, #fundings-exits, #glg, #justin-kan, #machine-learning, #mailchimp, #n26, #neo4j, #profounders, #real-time-data, #saas, #software-as-a-service, #tc, #travelperk, #twitch, #y-combinator, #zapier

0

Pinterest announces $500K Creator Fund, ‘Creator Code’ content policy, moderation tools and more

Pinterest today hosted an event focused on its creator community, where the company announced a series of updates including the launch of a $500,000 Creator Fund, a new content policy called the Creator Code, as well as new moderation tools, among other things. With the changes, the company says its goal is to ensure the platform continues to be a “inclusive, positive and inspiring place.” The new content guidelines put that into more specific terms as it requires Pinterest creators to fact-check content, practice inclusion, be kind, and ensure any call to action they make via the site doesn’t cause harm.

Creators will be required to agree and sign the code during the publishing process for Story Pins, where they tap a button that say “I agree” to statements that include “Be Kind,” “Check my facts,” “Be aware of triggers,” “Practice inclusion,” and “Do Not Harm.”

Image Credits: Pinterest

The code will be enforced the same way Pinterest today applies its rules for its other content policies: a combination of machine learning and human review, Pinterest tells us. However, the site’s algorithm will be designed to reward positive content and block harmful content, like anti-vaccination sentiments, for example. This could have a larger impact on what sort of content is shared on Pinterest, rather than a pop-up agreement with simple statements.

The Creator Code itself is not yet live, but will roll out to creators to sign and adopt in the weeks ahead, Pinterest says.

Image Credits: Pinterest

Pinterest today also introduced several new creator tools focused on the similar goal of making Pinterest a more positive, safe experience for all.

It’s launching comment moderation tools that will allow creators to remove and filter comments on their content, as well as tools that will allow them to feature up to three comments in the comment feed to highlight positive feedback. New spam prevention tools will help to clear out some of the unwanted comments, too, by leveraging machine learning technology to detect and remove bad comments.

Also new are “positivity reminders,” which will pop up asking Pinterest users to reconsider before posting potentially offensive comments. The notification will push users to go back and edit their comment, but doesn’t prevent them from posting.

Image Credits: Pinterest

Related to these efforts, Pinterest announced the launch of its first-ever Creator Fund at today’s event. The fund is specifically focused on elevating creators from underrepresented communities in the United States, and will offer a combination of creative strategy consulting, and compensating them with budget for content creation and ad credits. At least 50% of the fund’s recipients will be from underrepresented groups, Pinterest says.

The company tells us it’s initially committed to giving creators $500,000 in cash and media throughout 2021.

“For the first participants of the program, we worked with eight emerging creators across fashion, photography, food and travel, and will be identifying ten more creators in the next few months for the next cohort,” noted Creator Inclusion Lead Alexandra Nikolajev.

“We’re on a journey to build a globally inclusive platform where Pinners and Creators around the world can discover ideas that feel personalized, relevant and reflective of who they are,” Nikolajev said.

Pinterest has been working to rebuild its image in the wake of last year’s allegations of a host of internal issues, including unfair pay, racism, retaliation, and sexism, which conflicted with its outside image of being one of the “nicer” places to work in tech. Despite this fallout — which included a lawsuit, employee walkout, petitions, and more —  the issues that had been raised weren’t always reflected in Pinterest’s product.

The company had previously launched inclusive features like “skin tone ranges” to help those shopping for beauty products find matches for their skin tone. It also allowed retailers and brands to identify themselves as members of an underrepresented group, which gave their content the ability to appear in more places across Pinterest’s platform, like the Today tab, Shopping Spotlights and The Pinterest Shop, for instance.

Evan Sharp, Pinterest’s co-founder and Chief Design and Creative Officer, referenced the company’s image as “a positive place” at today’s event.

“We’ve been building Pinterest for 11 years, and ever since our users routinely tell us that Pinterest is the ‘last positive corner of the internet.’ In that time, we’ve also learned that you need to design positivity into online platforms as deliberately as much as you design negativity out,” Sharp said. “The Creator Code is a human-centric way for Creators to understand how to be successful on Pinterest while using their voice to keep Pinterest positive and inclusive,” he added.

Today, Pinterest serves over 450 million users worldwide, but is challenged by large platforms serving creators like Facebook, Instagram, YouTube, and others, including newcomers like TikTok and those that are inching into the creator community with funds of their own, like Snapchat, which is paying creators for Spotlight content, and Clubhouse, which is now funding creators’ shows. The increased competition for creator interest has left Pinterest needing an incentive program of its own.

To kick of its announcement, Pinterest’s Head of Content and Creator Partnerships, Aya Kanai, interviewed television personality Jonathan Van Ness (Queer Eye) at today’s virtual event, where they talked about the need for positivity and inclusivity on social media. Other event participants included creators Peter Som, Alison Cayne, Onyi Moss, Oyin Edogi and Jomely Breton — the latter two who spoke about putting the Creator Fund to use for themselves.

#apps, #creator-fund, #evan-sharp, #machine-learning, #machine-learning-technology, #online-platforms, #pinterest, #social, #social-media, #social-networks, #united-states

0

Streamlit nabs $35M Series B to expand machine learning platform

As a company founded by data scientists, Streamlit may be in a unique position to develop tooling to help companies build machine learning applications. For starters, it developed an open source project, but today the startup announced an expanded beta of a new commercial offering and $35 million Series B funding.

Sequoia led the investment with help from previous investors Gradient Ventures and GGV Capital. Today’s round brings the total raised to $62 million, according to the company.

Data scientists can download the open source project and build a machine learning application, but it requires a certain level of technical aptitude to make all the parts work. Company co-founder and CEO Adrien Treuille  says that so far the company has 20,000 monthly active developers using the open source tooling to develop streaming apps, which have been viewed millions of times.

As they have gained that traction, they have customers who would prefer to use a commercial service. “It’s great to have something free and that you can use instantly, but not every company is capable of bridging that into a commercial offering,” Treuille explained.

Company COO and co-founder Amanda Kelly says that the commercial offering called Streamlit for Teams is designed to remove some of the complexity around using the open source application. “The whole [process of] how do I actually deploy an app, put it in a container, make sure it scales, has the resources and is securely connected to data sources […] — that’s a whole different skill set. That’s a DevOps and IT skill set,” she said.

What Streamlit for Teams does is take care of all that in the background for end users, so they can concentrate on the app building part of the equation without help from the technical side of the company to deploy it.

Sonya Huang, a partner at Sequoia, who is leading the firm’s investment in Streamlit, says that she was impressed with the company’s developer focus and sees the new commercial offering as a way to expand usage of the applications that data scientists have been building in the open source project.

“Streamlit has a chance to define a better interface between data teams and business users by ushering in a new paradigm for interactive, data-rich applications,” Huang said.

They have data scientists at big-name companies like Uber, Delta Dental and John Deere using the open source product already. They have kept the company fairly lean with 27 employees up until now, but the plan is to double that number in the coming year with the new funding, Kelly says.

She says that the founding team recognizes that it’s important to build a diverse company. She admits that it’s not always easy to do in practice when as a young startup, you are just fighting to stay alive, but she says that the funding gives them the luxury to step back and begin to hire more deliberately.

“Literally right before this call, I was on with a consultant who is going to come in and work with the executive team, so that we’re all super clear about what we mean [when it comes to] diversity for us and how is this actually a really core part of our company, so that we can flow that into recruiting and people and engineering practices and and make that a lived value within our company,” she said.

Streamlit for Teams is available in beta starting today. The company plans to make it generally available some time later this year.

#developer, #funding, #machine-learning, #open-source, #recent-funding, #sequoia, #startups, #streamlit, #tc

0

Snorkel AI scores $35M Series B to automate data labeling in machine learning

One of the more tedious aspects of machine learning is providing a set of labels to teach the machine learning model what it needs to know. Snorkel AI wants to make it easier for subject matter experts to apply those labels programmatically, and today the startup announced a $35 million Series B.

It also announced a new tool called Applications Studio that provides a way to build common machine learning applications using templates and predefined components.

Lightspeed Venture Partners led the round with participation from previous investors Greylock, GV, In-Q-Tel and Nepenthe Capital. New investors Walden and BlackRock also joined in. The startup reports that it has now raised $50 million.

Company co-founder and CEO Alex Ratner says that data labeling remains a huge challenge and roadblock to moving machine learning and artificial intelligence forward inside a lot of industries because it is costly, labor-intensive and hard for the subject experts to carve out the time to do it.

“The not so hidden secret about AI today is that in spite of all the technological and tooling advancements, roughly 80 to 90% of the cost and time for an average AI project goes into just manually labeling and collecting and relabeling this training data,” he said.

He says that his company has developed a solution to simplify this process to make it easier for subject experts to programmatically add the labels, a process he says decreases the time and effort required to apply labels in a pretty dramatic way from months to hours or days, depending on the complexity of the data.

As the company has developed this methodology, customers have been asking for help in the next step of the machine learning process, which is taking that training data and the model and building an application. That’s where the Application Studio comes in. It could be a contract classifier at a bank or a network anomaly detector at a telco and it helps companies take that next step after data labeling.

“It’s not just about how you programmatically label the data, it’s also about the models, the preprocessors, the post processors, and so we’ve made this now accessible in a kind of templated and visual no-code interface,” he said.

The company’s products are based on research that began at the Stanford AI Lab in 2015. The founders spent four years in the research phase before launching Snorkel in 2019. Today, the startup has 40 employees. Ratner recognizes the issues that the technology industry has had from a diversity perspective and says he has made a conscious effort to build a diverse and inclusive company.

“What I can say is that we tried to prioritize it at a company level, the full team level and at a board level from day one, and to also put action behind that. So we’ve been working with external firms for internal training and audits and strategy around DEI, and we’ve made pipeline diversity, a non-negotiable requirement of any of our contracts with recruiting firms,” he said.

Ratner also recognizes that automation can hard code bias into machine learning models, and he’s hopeful that by simplifying the labeling process, it can make it much easier to detect bias when it happens.

“If you start with a dozen or two dozen of what we call labeling functions in Snorkel, you still need to be vigilant and proactive about trying to detect bias, but it’s easier to audit what taught your model to change it by just going back and looking at a couple of hundred lines of code.”

#artificial-intelligence, #data-labeling-tools, #funding, #lightspeed-venture-partners, #machine-learning, #recent-funding, #snorkel-ai, #startups, #tc

0

Hiro Capital puts $2.3M into team sports tracking platform PlayerData — as does Sir Terry Leahy

Hiro Capital has gradually been making a name for itself as an investor in the area know as ‘Digital Sports’ or DSports for shorts. It’s now led a $2.3m funding round in PlayerData. While the round might sound small, the area it’s going into is large and growing. Also investing in the round is Sir Terry Leahy, previously the CEO of Tesco, the largest British retailer.

Edinburgh, UK-based PlayerData uses wearable technology and software tracking to give grass-roots and professional sports teams feedback on their training. It can, for instance, allow coaches to replay key moments from a game, even modeling different outcomes based on player positioning.

This is Hiro Capital’s 4th DSports and ‘connected fitness’ investment, and it joins Zwift, FitXR and NURVV. Hiro has also invested in eight games startups in the UK, USA and Europe, as befits the heritage of cofounder and partner Ian Livingstone, OBE,CBE, who is the former chairman of Tomb Raider publisher Eidos plc and all-round gaming pioneer.

PlayerData says it has captured more than 10,000 team sessions across UK soccer and rugby, and logged over 50 million meters of play. It also has strong network effects, it says. Every time a new team encounters one using Playerdata’s platform, it generates 5 more clubs as users.

Roy Hotrabhvanon is cofounder and CEO of PlayerData, and is a former international-level archer. He’s joined by Hayden Ball, cofounder and CTO, a firmware and cloud infrastructure expert.

playerdata app

playerdata app

In a statement Hotrabhvanon said: “Our mission is to bring fine-grained data and insight to clubs across team sports, helping them supercharge their game-making, improve player performance, and avoid injury… Our ultimate goal is to implement cutting-edge insights from pioneering wearables that are applicable to any team in any discipline at any level.”

Cherry Freeman, co-founding Partner at Hiro says: “PlayerData ticks all of our key boxes: a huge TAM with over 3m grass-roots clubs; a deep moat built on shared player data, machine learning and highly actionable predictive algorithms; compelling customer network effects; and a really impressive yet humble founding team.”

The PlayerData news forms part of a wider growth in digital sports, which includes such breakout names as Peloton, Tonal, Mirror, as well as Hiro’s portfolio investment, Zwift. With the pandemic putting an emphasison both home workouts and general health, the fascination with digital measurement of performance now has a growing grip on the sector.

Speaking to TechCrunch, Freeman added: “We think there are something like 3 million teams that are potential customers for PlayerData. Obviously the number of runners is enormous, and they only need to get a small slice of that market to have a very, very large business. At the end of the day everyone, everyone works out, even if you just go for a walk, so the target market’s huge and they started with running but their technology is applicable to a whole raft of other sports.”

#capital, #ceo, #chairman, #cofounder, #cycling, #edinburgh, #europe, #fiction, #finance, #hiro, #machine-learning, #partner, #player, #tc, #tesco, #tonal, #united-kingdom, #united-states, #wearable-technology, #zwift

0

Note-taking app Mem raises $5.6 million from Andreessen Horowitz

The competition for note-taking is as fierce as it has ever been with plenty of highly-valued productivity startups fighting for an audience it can potentially serve endless productivity offshoots. In the past year, Notion raised at a $2 billion valuation, Coda raised at $636 million, and Roam raised at $200 million.

A new competitor in the space is emerging out of stealth with fresh funding from Andreessen Horowitz. The free app, called Mem, is an early access platform dedicated to pushing users to quickly jot down their thoughts without focusing too heavily on the underlying organization of them. The startup’s founders have vast ambitions for what their platform could become down the road, tapping into further advances in machine learning and even AR.

“Really the differentiation is [information] that is summonable ubiquitously wherever you are,” Mem co-founder Kevin Moody tells TechCrunch. “So, in the near term, through your desktop app with Mem Spotlight as a heads-up display for wherever you are, in the medium term through an assistive mobile application, and then in the long term, imagine contact lenses that are overlaying useful content to you in the world.”

Moody and his co-founder Dennis Xu tell TechCrunch they’ve raised $5.6 million led by a16z with additional participation from their Cultural Leadership Fund, Will Smith’s dreamers.vc, Floodgate and Unusual Ventures. The round also was host to a handful of angel investors including Harry Stebbings, Julia Lipton, Niv Dror, Tony Liu, Rahul Vohra and Todd Goldberg, among others.

In its current iteration, Mem push users towards “lightweight organization” rather than clicking through folders and links to find the perfect place to nestle their thoughts. Users can quickly tag users or dedicated topics in their notes. The user workflow relies pretty heavily on search and chronological organization, presenting users with their most recently accessed notes. Users can also set reminders for certain notes, bringing a popular email framework to note-taking.

For users of stock apps like Apple Notes, these interface quirks may not sound very jarring, though the design is still a departure from apps like Notion and Airtable which have heavily focused on structure over immediacy.

Mem Spotlight

Perhaps Mem’s biggest shift is how users access the information they’ve dumped into the platform. The founders say they want to avoid their app being seen as a “destination,” instead hoping users rely heavily on a keyboard-shortcut-prompted overlay called Mem Spotlight that allows them to search out information that they may need for an email, presentation or text message. The broader hope of the founders and investors behind Mem is that the team can leverage the platform’s intelligence over time to better understand the data dump from your brain — and likely other information sources across your digital footprint — to know you better than any ad network or social media graph does.

“What would it mean to just capture passively your digital footprint and then make use of that as though it were structured,” Moody posits. “If we can actually have our own Mem modeling of all of these entities, whether it’s text, or maybe it’s contacts, the people that you know, or it’s the events that you’re going to and these different sources feed into Mem, what would it mean for Mem to be able to have a product that is the ‘you’ API?”

For now, the startup’s app isn’t quite as grandiose in scale as what the founders may see in its future, but as Mem continues to onboard early users from its waitlist and add to its desktop functionality, the company is driving towards a platform they hope feels more instrumental to how its users “remember” information.

#ad-network, #andreessen, #andreessen-horowitz, #api, #co-founder, #harry-stebbings, #machine-learning, #notion, #rahul-vohra, #tc, #todd-goldberg, #unusual-ventures, #will-smith

0