Tech nabs $35M Series B as it releases machine learning feature store – TechCrunch is a startup founded by three former Uber engineers who wanted to share the idea of ​​a machine learning feature store with the public, just seven months after launching the $ 20 million Series A. Today we announced the $ 35 million Series B.

When we talked to the company in April, they were working with early customers in beta of the product, but today they are announcing general availability of the platform in addition to funding.

Like Series A, this round has Andreessen Horowitz And Sequoia Capital is back to jointly lead the investment. The company is currently raising $ 60 million.

The reason these two companies are so enthusiastic about Tecton is the specific machine learning problem they are trying to solve. “We help organizations bring machine learning into production. That’s the overall goal of our company, helping someone build operational machine learning applications. That means applications that enhance fraudulent systems or something realistic for them. […] And we’ll make it easy for them to build, deploy, and maintain, “explained Mike Del Balso, CEO and co-founder of the company.

They do this by providing the concept of a feature store, the ideas they came up with, and ideas that are themselves becoming a category of machine learning. Just last week, AWS announced the Sagemaker Feature store. This is what the company considers to be the primary validation of the idea.

As Tecton defines, a feature store is an end-to-end machine learning management system that contains a pipeline that transforms data into what is called a feature value, storing and managing all its feature data, and finally. A set of data that provides a consistent service.

Del Balso states that this works in conjunction with other layers of the machine learning stack. “When building a machine learning application, we use a machine learning stack that can contain layers like model training systems, model delivery systems, or MLOps that do all the model management. Then there is the feature management layer. We are a feature store — therefore we have an end-to-end life cycle of the data pipeline, ”he said.

With so much money behind the company, it’s growing rapidly, from 17 employees to 26 since we talked in April with plans to more than double that number by the end of next year. became. Del Balso says he and his co-founders are committed to building a diverse and comprehensive company, but admits that it’s not easy.

“This is, in fact, what we have a major recruitment initiative. It’s very difficult and takes a lot of effort. It allows you to make it like a second priority and make it It’s not something you can take seriously, “he said. To that end, he said, the company has sponsored and attended diversity recruitment meetings and focused its recruitment efforts on finding diverse candidates.

Unlike many startups we’ve talked to, Del Balso wants to get back to the office setup as soon as possible as a way to build more personal connections between employees. nabs $35M Series B as it releases machine learning feature store – TechCrunch Source link nabs $35M Series B as it releases machine learning feature store – TechCrunch

Related Articles

Back to top button