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

Union.ai, a startup emerging from stealth with a commercial version of the open source AI orchestration platform Flyte, announced today that it has raised $10 million in a round contributed by NEA and “select” angel investors. CEO Ketan Umare says proceeds will go to support the Flyte community by “improving Flyte’s accessibility, performance and reliability” and expanding the range of systems Flyte integrates with.

While businesses find the predictive power of AI appealing, particularly on the data analytics side of the organization, achieving meaningful results with AI often proves challenging. It is true that AI can help project revenue, for example by identifying trends in buying and selling. But implementing and maintaining the data pipelines necessary to prevent AI systems from drifting into inaccuracy can require substantial technical resources.

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

“[Union’s] the 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. “The search for the right solution led the team to delve into machine learning techniques, which came with requirements to use large amounts of data and consistently deliver robust models to production… The techniques used were based on platforms and the solution was widely used at Lyft.”

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

“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 infrastructure teams,” continued Umare. “Us [founded Union] because we believe that machine learning and data workflows are fundamentally different from software implementations. This is because software is more accurate with a slower lifecycle, while machine learning and data workflows start out experimental and may need to be put into production quickly.”

taking flotation

The other co-founders of Umare and Union, Haytham Abuelfutuh and George Snelling, have extensive experience in the technology industry. Prior to joining Lyft, Umare was a Senior Software Engineer at Amazon and Principal Engineer at Oracle, where he led the development of a block storage product for a bare metal and infrastructure-as-a-service offering. Abuelfutuh spent seven years as an engineer at Microsoft and three as a developer at Google, where he helped ship an internal software library for his own 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, which launches to coincide with Flyte’s version 1.0 release, Umare says the goal is to reduce (and ideally eliminate) the unwieldy infrastructure that can come with data science projects and muscle building. hamstrings In the worst case, messy abstractions can require rebuilding the infrastructure to implement AI in production, Umare says, negatively affecting the potential return on investment.

According to a 2021 Wakefield Research report, enterprise data engineers spend nearly half of their time building and maintaining data pipelines. Sixty-nine percent of survey respondents, primarily data engineers, said business results would improve if their teams could contribute more to business decisions and spend less time manually managing pipelines.

“Production machine learning is still in its infancy right now, especially in companies outside of big tech. Therefore, most companies start with DIY, that is our main competence”, said Umare. “We took a radically different first principles approach to defining what a workflow means for machine learning and data scientists. We started with the goal of minimizing human error and trying to help predict problems early. [and worked] closely with an extremely sophisticated and diverse set of partners such as Spotify, Gojek and Freenome [to help] refine the solution.”

Union Cloud inherits all the features and capabilities of Flyte, including connectors between compute backends that log all changes in an AI pipeline. Union Cloud also stores a history of all runs in a pipeline and provides a dashboard, command line interface, and API to interact with calculations.

Union Cloud and Flyte define workflows as multitasking. Workflows and tasks can be written in any programming language and remain on-premises, as can the data that moves through those components.

cloud advantage

So what is the added value with Union Cloud? Umare says 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 therefore our users can use [it] to build a self-service 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 available in the academic world, which is difficult to produce. We need to bring both worlds together 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 Flyte open source project cannot. While he acknowledges that there are similar efforts from other vendors, such as AWS Sagemaker, he believes that they fail to integrate well with the rest of the data science ecosystem.

“We’ve been on this problem for over five years, refining our solution and iterating based on real-world feedback and requirements,” said Umare. “The machine learning industry is already big and is growing within traditional companies as well. However, we believe that the potential for growth is not limited by the size of current demand, but rather by the experience we can offer, which is why we have focused solely 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 workforce by the end of the year as it focuses on product creation. Umare had no stats to share on interest or uptake of Union Cloud, but he reiterated that “thousands” of users at companies like Lyft, Spotify, Toyota subsidiary Woven Planet, and biotech and finance brands have adopted Flyte.

Leave a Comment