Personal blog of Yzmir Ramirez

Iterative, an open-source startup that is building an enterprise AI platform to help companies operationalize their models, today announced that it has raised a $20 million Series A round led by 468 Capital and Mesosphere co-founder Florian Leibert. Previous investors True Ventures and Afore Capital also participated in this round, which brings the company’s total funding to $25 million.

The core idea behind Iterative is to provide data scientists and data engineers with a platform that closely resembles a modern GitOps-driven development stack.

After spending time in academia, Iterative co-founder and CEO Dmitry Petrov joined Microsoft as a data scientist on the Bing team in 2013. He noted that the industry has changed quite a bit since then. While early on, the questions were about how to build machine learning models, today the problem is how to build predictable processes around machine learning, especially in large organizations with sizable teams. “How can we make the team productive, not the person? This is a new challenge for the entire industry,” he said.

Big companies (like Microsoft) were able to build their own proprietary tooling and processes to build their AI operations, Petrov noted, but that’s not an option for smaller companies.

Currently, Iterative’s stack consists of a couple of different components that sit on top of tools like GitLab and GitHub. These include DVC for running experiments and data and model versioning, CML, the company’s CI/CD platform for machine learning, and the company’s newest product, Studio, its SaaS platform for enabling collaboration between teams. Instead of reinventing the wheel, Iterative essentially provides data scientists who already use GitHub or GitLab to collaborate on their source code with a tool like DVC Studio that extends this to help them collaborate on data and metrics, too.

Image Credits: Iterative

“DVC Studio enables machine learning developers to run hundreds of experiments with full transparency, giving other developers in the organization the ability to collaborate fully in the process,” said Petrov. “The funding today will help us bring more innovative products and services into our ecosystem.”

Petrov stressed that he wants to build an ecosystem of tools, not a monolithic platform. When the company closed this current funding round about three months ago, Iterative had about 30 employees, many of whom were previously active in the open-source community around its projects. Today, that number is already closer to 60.

“Data, ML and AI are becoming an essential part of the industry and IT infrastructure,” said Leibert, general partner at 468 Capital. “Companies with great open-source adoption and bottom-up market strategy, like Iterative, are going to define the standards for AI tools and processes around building ML models.”

Directly, a startup whose mission is to help build better customer service chatbots by using experts in specific areas to train them, has raised more funding as it opens up a new front to grow its business: APIs and a partner ecosystem that can now also tap into its expert network. Today Directly is announcing that it has added $11 million to close out its Series B at $51 million (it raised $20 million back in January of this year, and another $20 million as part of the Series B back in 2018).

The funding is coming from Triangle Peak Partners and Toba Capital, while its previous investors in the round included strategic backers Samsung NEXT and Microsoft’s M12 Ventures (who are both customers, alongside companies like Airbnb), as well as Industry Ventures, True Ventures, Costanoa Ventures and Northgate. (As we reported when covering the initial close, Directly’s valuation at that time was at $110 million post-money, and so this would likely put it at $120 million or higher, given how the business has expanded.)

While chatbots have now been around for years, a key focus in the tech world has been how to help them work better, after initial efforts saw so many disappointing results that it was fair to ask whether they were even worth the trouble.

Directly’s premise is that the most important part of getting a chatbot to work well is to make sure that it’s trained correctly, and its approach to that is very practical: find experts both to troubleshoot questions and provide answers.

As we’ve described before, its platform helps businesses identify and reach out to “experts” in the business or product in question, collect knowledge from them, and then fold that into a company’s AI to help train it and answer questions more accurately. It also looks at data input and output into those AI systems to figure out what is working, and what is not, and how to fix that, too.

The information is typically collected by way of question-and-answer sessions. Directly compensates experts both for submitting information as well as to pay out royalties when their knowledge has been put to use, “just as you would in traditional copyright licensing in music,” its co-founder Antony Brydon explained to me earlier this year.

It can take as little as 100 experts, but potentially many more, to train a system, depending on how much the information needs to be updated over time. (Directly’s work for Xbox, for example, used 1,000 experts but has to date answered millions of questions.)

Directly’s pitch to customers is that building a better chatbot can help deflect more questions from actual live agents (and subsequently cut operational costs for a business). It claims that customer contacts can be reduced by up to 80%, with customer satisfaction by up to 20%, as a result.

What’s interesting is that now Directly sees an opportunity in expanding that expert ecosystem to a wider group of partners, some of which might have previously been seen as competitors. (Not unlike Amazon’s AI powering a multitude of other businesses, some of which might also be in the market of selling the same services that Amazon does).

The partner ecosystem, as Directly calls it, use APIs to link into Directly’s platform. Meya, Percept.ai, and SmartAction — which themselves provide a range of customer service automation tools — are three of the first users.

“The team at Directly have quickly proven to be trusted and invaluable partners,” said Erik Kalviainen, CEO at Meya, in a statement. “As a result of our collaboration, Meya is now able to take advantage of a whole new set of capabilities that will enable us to deliver automated solutions both faster and with higher resolution rates, without customers needing to deploy significant internal resources. That’s a powerful advantage at a time when scale and efficiency are key to any successful customer support operation.”

The prospect of a bigger business funnel beyond even what Directly was pulling in itself is likely what attracted the most recent investment.

“Directly has established itself as a true leader in helping customers thrive during these turbulent economic times,” said Tyler Peterson, Partner at Triangle Peak Partners, in a statement. “There is little doubt that automation will play a tremendous role in the future of customer support, but Directly is realizing that potential today. Their platform enables businesses to strike just the right balance between automation and human support, helping them adopt AI-powered solutions in a way that is practical, accessible, and demonstrably effective.”

In January, Mike de la Cruz, who took over as CEO at the time of the funding announcement, said the company was gearing up for a larger Series C in 2021. It’s not clear how and if that will be impacted by the current state of the world. But in the meantime, as more organizations are looking for ways to connect with customers outside of channels that might require people to physically visit stores, or for employees to sit in call centres, it presents a huge opportunity for companies like this one.

“At its core, our business is about helping customer support leaders resolve customer issues with the right mix of automation and human support,” said de la Cruz in a statement. “It’s one thing to deliver a great product today, but we’re committed to ensuring that our customers have the solutions they need over the long term. That means constantly investing in our platform and expanding our capabilities, so that we can keep up with the rapid pace of technological change and an unpredictable economic landscape. These new partnerships and this latest expansion of our recent funding round have positioned us to do just that. We’re excited to be collaborating with our new partners, and very thankful to all of our investors for their support.”

The technology that runs our companies these days is staggering in its complexity. We have moved from a monolith to a microservices world, from boxes to SaaS, and while that has added agility to the enterprise, it has come at the cost of a metric f-ton of services and software platforms required by every team in the building.

CIOs need a place to commiserate and get better recommendations on what tech works well and what should be placed in the proverbial recycle bin. Meanwhile, salespeople and investors want to hear these decision-makers’ views on emerging products to identify rich veins to invest in.

At the core of Pulse is a community of vetted CIOs and other tech procurers, currently numbering more than 15,000. On top of this core group of users, Pulse has built a series of products to help exploit their collective wisdom, including several new products the company is announcing today.

In addition to new product launches, the company is announcing a $6.5 million Series A round from AV8 Ventures, which is exclusively backed by mega-insurer Allianz Group and launched last year with a debut $170 million fund. This round closed in December according to the company and brings the startup’s total funding to $10.5 million.

Pulse’s existing product offerings assist product marketers and investment researchers who want to get a “pulse” on the marketplace for tech products by polling CIOs and testing out language around new features and initiatives.

“As an example, Microsoft will come to us and say, ‘Hey, we want to test our messaging and positioning before we sort of blow it up as a campaign. We’d like to do that very quickly through your community.’ And then we facilitate that through a series of questions through surveys and get back the insights to them very quickly,” co-founder and CEO Mayank Mehta explained.

“We think about this as truly becoming a Bloomberg terminal for marketers and investors,” he said. Researchers “can use this as a great way to get a real-time pulse on their buyers and understand how the market is moving, so they can make appropriate investments and ship strategies in real time.”

He said that the company worked with 50 customers last year and delivered some 150 reports. As for the CIOs themselves, “The community is open so long as you are a director level or above,” Mehta said.

In addition to this product for investors and market researchers, the company is also announcing the launch of Product IQ today, which takes the needs of a particular CIO user into account to offer them “personalized” product recommendations for their companies. Those recommendations are surfaced from the continuous data that CIOs are adding into the system through polls and opinion surveys.

“We’re trying to imagine and rethink how decision-making is done for technology executives, especially in a world like this where teams are changing so dramatically,” Mehta said.

Crowdsourced research platforms in the tech industry have become a popular area for VC investment in recent years. StackShare, which raised $5.2 million from e.Ventures, has focused on helping engineers learn from other engineers about the tech they have chosen for their infrastructure. Meanwhile, startups like Wonder and NewtonX, which raised $12 million from Two Sigma Ventures, have focused less on technical solutions and instead answer business questions such as market sizing or competitive landscape.

Pulse was founded in 2017 and is based in San Francisco, and previously raised a seed from True Ventures, according to Crunchbase.

Huge networks like Facebook and LinkedIn have a huge gravitational force in the world of social media — the size of their audiences make them important platforms for advertising and those who want information (for better or worse) to reach as many people as possible. But alongside their growth, we’re seeing a lasting role for platforms and networks focused on more narrow special interests, and today one of them — focused on farmers, of all communities — is picking up a round of funding to propel its growth.

WeFarm, a marketplace and networking site for small-holder farmers (that is, farms not controlled by large agribusinesses), has raised $13 million in a Series A round of funding, with plans to use the money to continue adding more users — farmers — and more services geared to their needs.

The round, which brings the total raised by the company to a modest $20 million, is being led by True Ventures, with AgFunder, June Fund; previous investors LocalGlobe, ADV and Norrsken Foundation; and others also participating.

WeFarm today has around 1.9 million registered users, and its early moves into providing a marketplace — helping to put farmers in touch with local suppliers of goods and gear such as seed and fertilizers — generated $1 million in sales in its first eight months of operations, a sign that there is business to be had here. The startup points out that this growth has been, in fact, “faster… than both Amazon and eBay in their early stages.”

WeFarm is based out of London, but while the startup does have users out of the U.K. and the rest of Europe, Kenny Ewan, the company’s founder and CEO, said in an interview that it is seeing much more robust activity and growth out of developing economies, where small-scale agriculture reigns supreme, but those working the farms have been massively underserved when it comes to new, digital services.

“We are building an ecosystem for global small-scale agriculture, on behalf of farmers,” Ewan said, noting that there are roughly 500 million small-scale farms globally, with some 1 billion people working those holdings, which typically extend 1.5-2 hectares and often are focused around staple commercial crops like rice, coffee, cattle or vegetables. “This is probably the biggest industry on Earth, accounting for some 75-80% of the global supply chain, and yet no one has built anything for them. This is significant on many levels.”

The service that WeFarm provides, in turn, is two-fold. The network, which is free to join, first of all serves as a sounding board, where farmers — who might live in a community with other farmers, but might also be quite solitary — can ask each other questions or get advice on agricultural or small-holding matters. Think less Facebook and more Stack Exchange here.

That provided a natural progression to WeFarm’s second utility track: a marketplace. Initially Ewan said that it’s been working with — and importantly, vetting — local suppliers to help them connect with farmers and the wider ecosystem for goods and services that they might need.

Longer term, the aim will be to provide a place where small-holding farmers might be able to exchange goods with each other, or sell on what they are producing.

In addition to providing access to goods for sale, WeFarm is helping to manage the e-commerce process behind it. For example, in regions like Africa, mobile wallets have become de facto bank accounts and proxies for payment cards, so one of the key ways that people can pay for items is via SMS.

“For 90% of our users, we are the only digital service they use, so we have to make sure we can fulfill their trust,” Ewan said. “This is a network of trust for the biggest industry on earth and we have to make sure it works well.”

For True and other investors, this is a long-term play, where financial returns might not be as obvious as moral ones.

“We are enormously inspired by how Kenny and the Wefarm team have empowered the world’s farmers, and we see great potential for their future,” said Jon Callaghan, co-founder of True Ventures, in a statement. “The company is not only impact-driven, but the impressive growth of the Wefarm Marketplace demonstrates exciting commercial opportunities that will connect those farmers to more of what they need to the benefit of all, across the food supply chain. This is a big, global business.”

Still, given the bigger size of the long tail, the company that can consolidate and manage that community potentially has a very valuable business on its hands, too.

Aurora Insight, a startup that provides a “dynamic” global map of wireless connectivity that it built and monitors in real time using AI combined with data from sensors on satellites, vehicles, buildings, aircraft and other objects, is emerging from stealth today with the launch of its first publicly available product, a platform providing insights on wireless signal and quality covering a range of wireless spectrum bands, offered as a cloud-based, data-as-a-service product.

“Our objective is to map the entire planet, charting the radio waves used for communications,” said Brian Mengwasser, the co-founder and CEO. “It’s a daunting task.” He said that to do this the company first “built a bunker” to test the system before rolling it out at scale.

With it, Aurora Insight is also announcing that it has raised $18 million in funding — an aggregate amount that reaches back to its founding in 2016 and covers both a seed round and Series A — from an impressive list of investors. Led by Alsop Louie Partners and True Ventures, backers also include Tippet Venture Partners, Revolution’s Rise of the Rest Seed Fund, Promus Ventures, Alumni Ventures Group, ValueStream Ventures and Intellectus Partners.

The area of measuring wireless spectrum and figuring out where it might not be working well (in order to fix it) may sound like an arcane area, but it’s a fairly essential one.

Mobile technology — specifically, new devices and the use of wireless networks to connect people, objects and services — continues to be the defining activity of our time, with more than 5 billion mobile users on the planet (out of 7.5 billion people) today and the proportion continuing to grow. With that, we’re seeing a big spike in mobile internet usage, too, with more than 5 billion people, and 25.2 billion objects, expected to be using mobile data by 2025, according to the GSMA.

The catch to all this is that wireless spectrum — which enables the operation of mobile services — is inherently finite and somewhat flaky in how its reliability is subject to interference. That in turn is creating a need for a better way of measuring how it is working, and how to fix it when it is not.

“Wireless spectrum is one of the most critical and valuable parts of the communications ecosystem worldwide,” said Rohit Sharma, partner at True Ventures and Aurora Insight board member, in a statement. “To date, it’s been a massive challenge to accurately measure and dynamically monitor the wireless spectrum in a way that enables the best use of this scarce commodity. Aurora’s proprietary approach gives businesses a unique way to analyze, predict, and rapidly enable the next-generation of wireless-enabled applications.”

If you follow the world of wireless technology and telcos, you’ll know that wireless network testing and measurement is an established field — about as old as the existence of wireless networks themselves (which says something about the general reliability of wireless networks). Aurora aims to disrupt this on a number of levels.

Mengwasser — who co-founded the company with Jennifer Alvarez, the CTO who you can see presenting on the company here — tells me that a lot of the traditional testing and measurement has been geared at telecoms operators, who own the radio towers, and tend to focus on more narrow bands of spectrum and technologies.

The rise of 5G and other wireless technologies, however, has come with a completely new playing field and set of challenges from the industry.

Essentially, we are now in a market where there are a number of different technologies coexisting — alongside 5G we have earlier network technologies (4G, LTE, Wi-Fi); and a potential set of new technologies. And we have a new breed of companies building services that need to have close knowledge of how networks are working to make sure they remain up and reliable.

Mengwasser said Aurora is currently one of the few trying to tackle this opportunity by developing a network that is measuring multiples kinds of spectrum simultaneously, and aims to provide that information not just to telcos (some of which have been working with Aurora while still in stealth) but the others kinds of application and service developers that are building businesses based on those new networks.

“There is a pretty big difference between us and performance measurement, which typically operates from the back of a phone and tells you when have a phone in a particular location,” he said. “We care about more than this, more than just homes, but all smart devices. Eventually, everything will be connected to network, so we are aiming to provide intelligence on that.”

One example are drone operators that are building delivery networks: Aurora has been working with at least one while in stealth to help develop a service, Mengwasser said, although he declined to say which one. (He also, incidentally, specifically declined to say whether the company had talked with Amazon.)

5G is a particularly tricky area of mobile network spectrum and services to monitor and tackle, which is one reason why Aurora Insight has caught the attention of investors.

“The reality of massive MIMO beamforming, high frequencies, and dynamic access techniques employed by 5G networks means it’s both more difficult and more important to quantify the radio spectrum,” said Gilman Louie of Alsop Louie Partners, in a statement. “Having the accurate and near-real-time feedback on the radio spectrum that Aurora’s technology offers could be the difference between building a 5G network right the first time, or having to build it twice.” Louie is also sitting on the board of the startup.

Over the past few years, the old language of “customer support” has been supplanted by the new language of “customer success.” In the old model, companies would essentially disappear following the conclusion of a sale, merely handling customer problems when they arose. Now, companies are actively reaching out to customers, engaging them with education and training and monitoring them with analytics to ensure they have the best time with the product as possible.

What’s changing is the nature of product and services today: subscription. Customers no longer just make a single buying decision about a product, but instead must actively commit to using the product, or else they churn.

New York-based Catalyst, founded by brothers Edward and Kevin Chiu, wants to rebuild customer success from the ground up with an integrated software platform. They have received some capital success of their own, securing $2.4 million in venture capital from Phil Black of True Ventures with participation from Ludlow Ventures and Compound.

New York has had something of an increase in founder mafias, as TechCrunch reported this weekend. Catalyst is no exception to this trend, with the Chiu brothers both working at DigitalOcean, one of New York’s many high-flying enterprise startups. Edward Chiu was director of customer success at the company for a number of years, but had a unique background in sales and also in coding before starting.

Kevin Chiu was head of inside sales at DigitalOcean . “I brought my brother on to do sales at DigitalOcean,” Edward Chiu explains. “We always knew that we wanted to start a company together, but wanted to see if we would kill each other.” The two worked together, and lo and behold, they didn’t kill each other.

Edward Chiu wanted to match the product experience of using DigitalOcean with the experience of using its internal customer success tools. Nothing on the market fit. “Given that DigitalOcean was a very technical product,” Chiu explained, “we decided to build our own tool.” Chiu thought of customer success at DigitalOcean as its own product, and his team built up the platform to improve its functionality and scalability. “We just used the tool and we loved it,” he said, so we “started to show this tool to a bunch of other customer success leaders I am connected with.”

Other customer success leaders said they wanted the platform, and “after the 20th person told me that,” he and his brother spun out of DigitalOcean to go on their own. Unlike enterprise startups in New York a couple of years ago that often struggled to find any investors, Catalyst found cash quickly. “Two weeks in we had more offers than we knew what to do with,” Chiu explained. The two said they had originally targeted a fundraise of $750,000, but ended up at $2.4 million.

Catalyst is a platform that integrates between a number of other major SaaS services such as Salesforce, Zendesk, Mixpanel and others to create a unified dashboard for data around customer success. From there, customer success managers have a set of automated tools to handle engagement, such as customer segmentation and email campaigns.

A major challenge in the customer success world is that these managers often don’t have the skills required to do advanced data analytics, so they often rely on their friends in engineering to run scripts or perform database lookups. The hope is that Catalyst’s feature set is powerful enough that these sorts of ad hoc tasks become a thing of the past. “Because we aggregate all this data, you can run queries,” Chiu explains.

Chiu says that Catalyst doesn’t just want to be a software platform, but rather a movement that pushes every company to think about how they can make their customers successful. “There are so many companies that are starting to understand that it is not something that you do once you raise a Series A, but something you do from day one,” Chiu said. “If you take care of your very first customer, they will constantly promote you and constantly promote your business.”

The company is based in Flatiron, and has eight employees.

 A couple of weeks ago, LinkedIn announced a new version of its recruitment product, giving HR teams the ability to look for new hires that most closely match the profiles of employees they already know and like. But LinkedIn was not the first to use search technology, machine learning and so-called “entity recognition” to update and improve the hiring process. A much… Read More