Union.ai raises $10M to simplify AI and ML workflow orchestration

Union.ai, a startup emerging from stealth with a commercial version of the open source AI orchestration platform Flyte, today announced that it raised $10 million in a round contributed by NEA and “select” angel investors. CEO Ketan Umare says that the proceeds will be put toward supporting the Flyte community by “improving the accessibility, performance and reliability of Flyte” and broadening the array of systems that Flyte integrates with.

While companies find AI’s predictive power alluring, particularly on the data analytics side of the organization, achieving meaningful results with AI often proves to be a challenge. It’s true that AI can help to project revenue, for example, by identifying trends in buying and selling. But implementing and maintaining the data pipelines necessary to keep AI systems from drifting to inaccuracy can require substantial technical resources.

That’s where Flyte comes in — a platform for programming and processing concurrent AI and data analytics workflows. Union’s team, including Umare, helped to build Flyte while at Lyft, where it was used to help create a system to calculate the estimated time of arrival (ETA) for drivers to get from point A to point B.

“[Union’s] founders first met at Lyft, where we joined the team responsible for calculating the ETA for a Lyft driver to get from point A to point B,” Umare told TechCrunch via email. “Searching for the right solution led the team deep into machine learning techniques, which came with requirements to use large amounts of data and deliver robust models to production consistently … The techniques used were platformized, and the solution was used widely at Lyft.”

Lyft contributed Flyte to open source in 2020, granting the trademark to the Linux Foundation a year later. That’s when Union’s team saw an opportunity to layer paid services on top of the project in the cloud.

“A managed version of Flyte, called Union Cloud, will allow smaller teams and organizations to use the power of Flyte without the need to staff up on infrastructure teams,” Umare continued. “We [founded Union] because we believe that machine learning and data workflows are fundamentally different from software deployments. This is because software is more precise with a slower lifecycle while machine learning and data workflows start off being experimental and may need to be quickly productionized.”

Taking Flyte

Umare and Union’s other co-founders, Haytham Abuelfutuh and George Snelling, all have deep backgrounds in the tech industry. Prior to joining Lyft, Umare was a senior software engineer at Amazon and a principal engineer at Oracle, where he led development of a block storage product for an infrastructure-as-a-service and bare metal offering. Abuelfutuh spent seven years as an engineer at Microsoft and three as a developer at Google, where he helped to ship an internal software library for first-party apps including Google Photos. Snelling — also a Microsoft veteran — co-founded several startups (Westside, LabKey and Patchr) and spent time at Salesforce as a senior director of engineering.

With Union Cloud — the launch of which coincides with the release of Flyte version 1.0 — Umare says the goal is to reduce (and ideally eliminate) the unwieldy infrastructure that can crop up in data science projects and hamstring development. At their worst, messy abstractions can necessitate rebuilding infrastructure to deploy AI to production, Umare points out — negatively affecting the potential return on investment.

According to a 2021 Wakefield Research report, enterprise data engineers spend nearly half their time building and maintaining data pipelines. Sixty-nine percent of respondents to the survey — mainly data engineers — said that business outcomes would improve if their teams could contribute more to business decisions and spend less time on manual pipeline management.

“Production machine learning is still in its infancy at the moment, especially at companies outside big tech. Thus, most companies start off with DIY — that is our primary competition,” Umare said. “We took a radically different, first-principles approach to defining what a workflow means for machine learning and data scientists. We started with a goal to minimize human errors and try to help predict problems ahead of time [and worked] closely with extremely sophisticated and a diverse set of partners like Spotify, Gojek and Freenome [to help] refine the solution.”

Union Cloud inherits all of Flyte’s characteristics and capabilities, including connectors between computation back ends that record all changes to an AI pipeline. Union Cloud also stores a history of all a pipeline’s executions and provides a dashboard, command-line interface and API to interact with the computations.

Union Cloud — and Flyte — define workflows as multiple tasks. Workflows and tasks can be written in any programming language and stay on-premises, as does data moving through those components.

Cloud advantage

So what’s the value add with Union Cloud? Umare says that it adds “agility, reproducibility, and security” to Flyte by centralizing infrastructure management and maintaining “high” privacy and compliance standards. “Our products are built with zero-trust principles in mind and thus our users can use [it] to build a self-serve platform that still maintains high security standards,” he continued. “Data science is very academic, which directly affects machine learning. There is a lot of fantastic research and literature that is available in academia, which is hard to productionize. We need to bridge both these worlds in a structured and repeatable way.”

Umare also sees Union Cloud as a way to reduce the cost of developing new products and systems in a way that the open source Flyte project can’t accomplish. While he concedes that similar efforts from other vendors exist, like AWS Sagemaker, he believes that they fail to integrate well with the rest of the data science ecosystem.

“We have been at this problem for over five years, refining our solution and iterating based on real-world feedback and requirements,” Umare said. “The machine learning sector is already large and growing within traditional companies as well. We view growth potential to not be limited by the size of the current demand however, but rather by the experience we can deliver, which is why we’ve focused purely on customer success and open source adoption. This will lead to revenue growth in the near future.”

On the topic of growth, Union plans to double its 20-person headcount by the end of the year as it focuses on product buildout. Umare didn’t have statistics to share on Union Cloud interest or uptake, but reiterated that “thousands” of users across companies such as Lyft, Spotify, Toyota subsidiary Woven Planet, and biotech and finance brands have adopted Flyte.

Union.ai, a startup emerging from stealth with a commercial version of the open source AI orchestration platform Flyte, today announced that it raised $10 million in a round contributed by NEA and “select” angel investors. CEO Ketan Umare says that the proceeds will be put toward supporting the Flyte community by “improving the accessibility, performance

Register for free and explore data trends at Data & the Culture Transformation

Data — it’s the lifeblood of every company, and as our data culture transforms, so too must your business. An emerging, innovative data ecosystem promises to integrate disparate data sources, provide a more complete picture of business, both present and future, and improve enterprise collaboration at every level.

Understanding the rapidly changing trends in data and analytics is more essential than ever, and it’s why we’re proud to host Data & the Culture Transformation, an online event — presented by Cloudera — on April 26 or 27, 2022.

Register today: Attending Data & the Culture Transformation is free, but you must register here to reserve your seat at the virtual table.

Pick your date and international showtimes:
April 26: Americas, EMEA and India
April 27: APAC and Singapore

Data & the Culture Transformation is tailor-made for anyone interested in building a data-driven future and deriving more value from their data — faster and more easily — to drive business success. We’re talking IT professionals, line-of-business leaders, data practitioners and decision makers.

What can you expect at this event? Tech industry analyst Maribel Lopez, of Lopez Research, will moderate conversations with Cloudera CTO Ram Venkatesh and Shirley Collie, chief health analytics actuary at Discovery Health.

Venkatesh will speak to the evolution in business culture and mindset when it comes to data, and how a more collaborative approach helps modern data-driven organizations get easier access to their data — and the insights it provides — from a variety of sources.

Tune in to hear Collie talk about how the South African health insurance company’s data-driven approach has benefited both their business and society.

These three diverse data experts will also sit down together and discuss a variety of topics, including how companies can respond to the current data explosion, the role ethics plays in data analytics and projections on the next data-driven culture transformation.

No matter what role you play in your organization’s data strategy, you’ll leave this symposium with actionable insights you can use now, when you need them most.

Attending Data & the Culture Transformation is 100% free. Register today, join us online and gain a deeper understanding of the changes driving new, collaborative data ecosystems — and how they can benefit your business.

Data — it’s the lifeblood of every company, and as our data culture transforms, so too must your business. An emerging, innovative data ecosystem promises to integrate disparate data sources, provide a more complete picture of business, both present and future, and improve enterprise collaboration at every level. Understanding the rapidly changing trends in data

Using AI to expand the quality and fairness of urban data

The sparse and inconsistent availability of urban data is currently hampering efforts to manage our cities fairly and effectively—but this could be solved by exploiting the latest advances in artificial intelligence.The sparse and inconsistent availability of urban data is currently hampering efforts to manage our cities fairly and effectively—but this could be solved by exploiting the latest advances in artificial intelligence.Machine learning & AI

How to build brain-inspired neural networks based on lighton April 12, 2022 at 9:21 am

Supercomputers are extremely fast, but also use a lot of power. Neuromorphic computing, which takes our brain as a model to build fast and energy-efficient computers, can offer a viable and much-needed alternative. The technology has a wealth of opportunities, for example in autonomous driving, interpreting medical images, edge AI or long-haul optical communications. Electrical engineer Patty Stabile is a pioneer when it comes to exploring new brain- and biology-inspired computing paradigms. “TU/e combines all it takes to demonstrate the possibilities of photon-based neuromorphic computing for AI applications.”Supercomputers are extremely fast, but also use a lot of power. Neuromorphic computing, which takes our brain as a model to build fast and energy-efficient computers, can offer a viable and much-needed alternative. The technology has a wealth of opportunities, for example in autonomous driving, interpreting medical images, edge AI or long-haul optical communications. Electrical engineer Patty Stabile is a pioneer when it comes to exploring new brain- and biology-inspired computing paradigms. “TU/e combines all it takes to demonstrate the possibilities of photon-based neuromorphic computing for AI applications.”

Jobox is building a bridge between skilled labor, available work

The global pandemic shifted many home service employees from W2 to 1099 status, which is giving Jobox.ai a lot more business.

The Florida and Bay Area-based company is going after the $595 billion home services industry, where skilled labor is facing a shortage, by connecting companies with over 5,000 vetted home service professionals via a marketplace leveraging an artificial intelligence-based infrastructure — thus eliminating the need for customers to vet professionals themselves.

Co-founders Shay Bloch, CEO, Kaushik Pendurthi, COO and Moshe Levi, who used to be a locksmith, started Jobox in 2016. One of the inspirations for the company was Bloch’s mother, who owned a restaurant for over 20 years.

“Even if you ask her today, what is the cost of sales and what is the bottom line of her business, and she won’t know. She says she just knows how to make good food and hope she has good service,” Bloch told TechCrunch. “I always knew that we can help with software for small businesses where we can create an impact, bring change to the bottom line and really change the livelihood in the industry that we chose.”

On one side, the home service professionals have a free toolbox right on their phone to get their small business up-and-running. On the other side, Jobox works with more than 50 large organizations that often outsource repair work to match jobs they have with the bank of professionals based on criteria, like skill set and location.

Jobox makes money from the transaction when the professional completes the job. For example, the professional gets 60% and Jobox and the originating organization split the difference.

Leveraging technology for the home repair market has become an attractive area for venture capital funding. Earlier this year, Zuper raised $13 million to help technicians make sure they had everything they needed when they showed up to a repair.

We also saw Fuzey bring in $4.5 million in seed funding for its “digital one-stop shop” for small businesses and independent contractors, and Puls Technologies get $15 million for its mobile app connecting tradespeople with on-demand home repair services. There are also larger companies in the space, like Jobber, which announced $60 million in funding.

Meanwhile, Jobox works with professionals in 39 states and has so far processed more than $1 billion of transactions on its platform. It did over $300 million last year alone in gross merchandise volume and $10 million in revenue, Bloch said.

It too attracted funding over the years — $58 million in total, and Tuesday marks its official “coming out of stealth mode” with a new $42 million of Series B funding. General Catalyst led the round and was joined by new and existing investors, including Resolute VC, NNS, Expanding Capital and Joey Low.

The new funding will enable the company to go into more cities, attract more professionals and expand into other home service positions — currently it is mainly locksmiths, garage door repair, pavement repair and carpet cleaning. They are looking at plumbing and electricians, too.

“We spent some time getting the product really right and making sure that our product is solving the right problems and the critical problems for our user base,” Pendurthi said. “We want to scale, and we want to have top talent. We also want to make sure that we’re building partnerships with different kinds of B2B companies so that we can serve our users even better, and we want to expand our products.”

Jobox connects companies with vetted professionals for home service jobs via a marketplace, thus eliminating the need for customers to vet professionals themselves.

Product analytics startup Kubit lands $18M in fresh capital

Showing that the product analytics industry is alive and well, Kubit today announced that it raised $18 million in a Series A funding round led by Insight Partners, bringing its total capital raised to $24 million. Kubit says that it’ll put the fresh cash toward growing its team and expanding its platform, which helps customers manage their data quality.

CEO Alex Li says that he founded Kubit in 2018 to solve what he believes is one of the biggest pain points of the product analytics space: losing data control and lack of transparency. Previously, Li spent 10 years building mobile apps and used product analytics tools at companies including Smule, Booyah and eBay. These tools fell short of his expectations, he says, in that they often required sending data to third parties and created siloed analytics practices.

“Product analytics has proven its significance in many large enterprises’ successes. [But as] an industry, product analytics is still young and has become more open and transparent,” Li told TechCrunch via email. “No enterprise wants to be locked into a siloed blackbox solution that demands control of a customer’s precious data.”

As a refresher, product analytics is the process of analyzing how users engage with a product — whether an app, website or subscription service — to allow companies to track and analyze user engagement data. The idea is to use the data to improve and optimize the product, and identify bugs as they crop up in the deployment process.

Kubit’s platform is designed to work with existing cloud data warehouses (i.e. central repositories of business data) without the need to transform, normalize or convert product analytics data. In this way, Kubit eliminates the need for batch jobs and the duplication of data, Li argues, potentially reducing cloud computing costs.

“We help our customers to analyze their users’ behavior patterns through their user data. Kubit doesn’t store or process any personally identifiable information and most of the user data is already anonymized,” Li said. “With our integration[s], our customers have full control of their data and can make changes whenever they want, including deletion.”

Li sees Amplitude, June and Mixpanel as direct competitors in the product analytics segment. Mixpanel and Amplitude have formidable warchests. But Li tells TechCrunch that both Kubit’s revenue and headcount (13 people) is projected to triple this year, driven by a growing customer base that includes “several largest enterprises in entertainment, social and education fields.”

“[I]t seems that the consumer segment has recovered very well [from the pandemic] and the surviving enterprises are the winners who really see the needs of product analytics to keep their growth going strong,” Li said. “Moreover, the down time [during the pandemic] gave many data teams the opportunity to reinforce their modern data stack and realize the significance of data control and transparency.”

Showing that the product analytics industry is alive and well, Kubit today announced that it raised $18 million in a Series A funding round led by Insight Partners, bringing its total capital raised to $24 million. Kubit says that it’ll put the fresh cash toward growing its team and expanding its platform, which helps customers

PassiveLogic, which creates digital twins of building systems, raises $15M

PassiveLogic, a startup developing a platform to autonomously control building systems, today announced that it raised $15 million from Brookfield Growth, the investment arm of asset management firm Brookfield. CEO Troy Harvey says that the new capital will be put toward growing PassiveLogic’s team and “launching an ecosystem of products that enable autonomy.”

Analysts at McKinsey (among others) predict the pandemic will spur an interest in more comfortable, sustainable working spaces. But assuming that comes to pass, legacy industrial building controls threaten to make such projects daunting. According to market research agency ARC Advisory Group, there are roughly $65 billion worth of distributed building control systems nearing their end of life, with many of those systems over 25 years old.

“Right now, building controls utilize very limited information to make sure the building works accordingly,” Harvey told TechCrunch via email. “To get to the future of real estate, there needs to be a digital platform that can aggregate building data, [allowing] building managers to customize automation controls and act upon it in real-time.”

PassiveLogic provides this solution, Harvey claims, built around what he calls the “Hive” controller. PassiveLogic’s product is designed to enable autonomous control of legacy building systems by interfacing with them using a combination of sensors, equipment and devices that don’t need cloud connectivity.

Harvey founded Salt Lake City, Utah-based PassiveLogic in 2016 with Jeremy Fillingim. Harvey was previously the CEO of Heliocentric, an engineering firm that worked with clients to architect “next-generation” buildings. Fillingim was a partner at Mote Systems, where he designed a touchscreen universal remote control.

PassiveLogic hosts a software environment, Autonomy Studio, where customers can create building system models from CAD or 3D models. The software uses these models to generate a “physics-based digital twin” that conforms to a descriptive standard called Quantum. Applications written in Quantum can be deployed within PassiveLogic’s control hardware — the aforementioned Hive.

Harvey asserts that Quantum provides “virtual analogs” to real-world objects via algorithms that attempt to understand how a building’s equipment and systems interact. Based on these predictions, PassiveLogic makes control and management decisions for maintenance and operation.

“PassiveLogic is creating [a] platform for generalized autonomy. The aim is to empower anyone to effortlessly design their own custom applications — without requiring a team of Ph.D.s,” Harvey said. “[There’s] no programming required to understand how a building’s equipment and systems interact.”

Digital twin technologies, which digitally models real-world systems, aren’t new. GE, AWS and other companies offer products that allow customers to model digital twins of machines. London-based SenSat creates digital twin models of locations for construction, mining and energy projects. Meanwhile, startups like Lacuna and Nexar are building digital twins of entire cities.

But digital twin technologies share the same limitations, chief among them inaccurate modeling in the presence of inaccurate data. Indeed, the models are only as good as the data that’s used to develop them. As Gartner notes in a report: “It is difficult to anticipate the nature of the simulation models, data types, and data analysis of sensor data that might be necessary to support the design, introduction, and service life of the digital twins’ physical counterparts. While 3D geometry is sufficient to communicate the digital twin visually and how parts fit together, the geometric model may not be able to perform simulations of the behavior of the physical counterpart in use or operation. At the same time, the geometric model may not be able to analyze data if it is not enriched with additional information.”

Another potential pitfall is the proprietary nature of digital twin platforms. Digital twins with long life cycles, like those of building systems, might extend well beyond the lifespans of the software and hardware used to create and maintain them.

Harvey says that PassiveLogic’s Hive hardware is engineered to combat data inaccuracy, and he challenges the idea that PassiveLogic will ever leave its customers without the necessary software updates.

PassiveLogic, a startup developing a platform to autonomously control building systems, today announced that it raised $15 million from Brookfield Growth, the investment arm of asset management firm Brookfield. CEO Troy Harvey says that the new capital will be put toward growing PassiveLogic’s team and “launching an ecosystem of products that enable autonomy.” Analysts at

EvolutionIQ secures $21M to streamline insurance claims processing with AI

Processing claims at scale presents a challenge for insurers, particularly where the claims entail factors like complex underlying health conditions. According to data from the National Association of Insurance Commissioners, the second-most common complaint that insurance customers made in 2021 was claim delays, ranked only after unsatisfactory settlement offers.

The pandemic placed an additional strain on insurers, with an RGA survey finding that claims acceptance rates for permanent disability, critical illness and long-term care have been minimal over the past two years. Even with reduced claims requirements, the average end-to-end time for claims rose from 34 days pre-pandemic to 43 days.

A growing cohort of startups, including Alan, Tractable and Snapsheet, offers tools to help customers navigate the insurance claims process. But former Google AI tech leader Tomas Vykruta is taking a different tack with EvolutionIQ, which works with insurers to analyze claimant data and third-party information to identify “high-opportunity” claims — specifically those involving bodily injury.

EvolutionIQ today announced that it raised $21 million in a Series A round led by Brewer Lane Ventures with participation from FirstRound Capital, FirstMark Capital, Foundation Capital, Altai Ventures, Asymmetric Ventures and insurance carriers Reliance Standard Life, New York Life Ventures, Guardian Life and Sedgwick. It brings the company’s total capital raised to $26.1 million at a “north of” $150 million valuation, following small seed and venture rounds in 2019 and 2020.

“EvolutionIQ assists insurance professionals in improving claims handling through insights uncovered by analyzing historical claims data,” Vykruta told TechCrunch via email. “With our decision intelligence platform, claims teams can recoup lost time and streamline processes. Our software allows front line operators to make more informed decisions and focus their energy on high-potential claims. For managers, we’re able to identify claims blocks that need further investigation and those that are easily resolved — and then provide guidance to make it happen.”

Accelerating insurance claims

Vykruta has a long and fascinating career in the tech industry. He’s the founder of two video game studios, Antartica and Electrolab, and was one of the lead console software architects at Surreal Software, a Warner Bros. Interactive Entertainment subsidiary. In 2008, Vykruta took a job at Microsoft as a senior engineer at the advanced technology group, where he worked on software for the Xbox 360. And in 2016, he joined Waymo, Google parent company Alphabet’s autonomous car division, as a machine learning engineer.

Vykruta spent the last roughly decade of his career at various Google machine learning teams prior to co-founding EvolutionIQ with Jonathan Lewin and Michael Saltzman, one of the founders of on-campus furniture rental business Roomie.

“While adoption of AI in a traditional industry can be challenging, [it can be] overcome by building ‘explainable AI’ systems in tight collaboration with the users,” Vykruta said. “There [must be] two systems: one that makes predictions [and] one that explains the forecasted outcomes in [plain] language.”

With EvolutionIQ, Vykruta, Lewin and Saltzman set out to design a platform that evaluates the history of claims — specifically health-related claims — up until today to answer questions like “Is this a claim on which we can take action?” and “Is there going to be an outcome that makes sense?” Using EvolutionIQ, Vykruta claims that a claims adjuster, who may have hundreds of cases active at a time, can better understand short-term disability, long-term disability, workers comp and property and casualty claims; identify the most actionable opportunities; and see the results in the context of new data and events.

“[C]laims are really incredibly complicated because you’ve got these narratives that are open for many years. Some of these claims are 15 or 20 years old. They have as many as 30 to 40 medical diagnoses. The insurance industry can’t solve this comprehensively,” Vykruta explained. “What this has led to is just incredible waste in unnecessary payments. Examiners [may] look at the last couple of pages of notes and generally, they don’t take any action at all.”

Informed by its predictive algorithms, EvolutionIQ spotlights dozens of claims from tens of thousands that are most likely to have the greatest outcome for claimants, carriers and clients. EvolutionIQ also monitors open claims to guide workers to those that require more attention or new actions, including claims that have potentially fallen through the cracks.

“Claims examiners want to focus on the cases that have the greatest impact on customers and carriers. But claims handling involves archaic and manual processes that require examiners to review too much information or to evaluate data on their own, despite the fact that each claim involves multiple people and systems,” Vykruta said. “Therefore, we built a decision intelligence platform that acts as an AI-enabled copilot to identify high-opportunity claims early in the lifecycle and cases most likely to be referred to adjusters.”

Market opportunity

The way Vykruta sees it, EvolutionIQ’s value proposition is reducing overpayment, waste and longer claim durations. Insurance carriers’ teams are stretched thin by rising claim volumes, and the systems responsible for making management scaleable actually compound the problem, he asserts — forcing examiners to make tradeoffs that lead to slower decisions and mistakes.

“Claims can last years, many worth six figures, and involve hundreds of pages and many formats (structured and unstructured) constantly flowing in … We actually give time back to claims organizations by identifying the cases that don’t need to be addressed and can be set to be resolved,” Vykruta said. “[We’re working with] the top disability carriers, property-casualty carriers and third-party administrators including Reliance Standard, Principal, Sun Life, Argo Group, Matrix Absence Management and FullscopeRMS.”

Of course, AI has a well-known bias problem, and reporting throughout the years has exposed how supposedly “fair” algorithms in insurance can perpetuate different forms of discrimination. Algorithms have led insurance companies to charge minority communities higher premiums than white communities, for example — even when the risks are the same. Asked how EvolutionIQ mitigates the potential for this sort of bias, Vykruta says that the company invests in “explainability” approaches that make it clear what sort of factors lead to the platform’s decisions.

“EvolutionIQ is not just taking the raw weight from the model. They translate the system that makes sense to [claims adjusters],” Vykruta said. “[The platform is] only showing them [items] that are actionable.”

Risks aside, EvolutionIQ isn’t alone in tackling the opportunity. Among its rivals are CCC Intelligent Solutions, a technology solutions provider for the automotive and insurance industries that relies on data science to expedite claims processing. Riskcovry is another startup in the nascent space, with customers including banks, fintechs and supply chain brands.

But the insurtech market is flush with cash, with one report estimating that global investments in insurance tech startups in 2021 topped $10 billion, up from $3.1 billion in 2020.

Vykruta says that EvolutionIQ — which was cash flow-positive in 2021 — will spend the bulk of the new capital on expanding its engineering, data science, product and customer success teams. Customer acquisition will also be a focus going forward, he says, as EvolutionIQ explores new and emerging categories of insurance.

Processing claims at scale presents a challenge for insurers, particularly where the claims entail factors like complex underlying health conditions. According to data from the National Association of Insurance Commissioners, the second-most common complaint that insurance customers made in 2021 was claim delays, ranked only after unsatisfactory settlement offers. The pandemic placed an additional strain