Slai makes smart decisions around machine learning setups for developers – TechCrunch


In a world where Twilio exists, you wouldn’t dream of building your own SMS messaging stack across 193 countries and God knows how many telecom operators. The situation for machine learning (ML) is not entirely different; If ML isn’t at the core of your software—and it probably isn’t—why waste time building an entire infrastructure. To solve exactly this problem, slay builds a developer-first platform for machine learning. It equips developers with the tools to quickly deploy machine learning applications.

“Today, machine learning remains a research discipline, and it’s still very difficult for a developer to create a custom machine learning application,” shares Eli Mernit, co-founder and CEO of Slai. “We hope that developers will be able to create state-of-the-art machine learning models.”

The company announced today that it has raised a $3.5 million seed round led by Tiger Global with additional investments from Y Combinator, Charge Ventures, Uncorrelated Ventures, Twenty Two Ventures and Soma Capital, as well as angels like Guy Podjarny and Jason Warner.

The company’s product focuses on allowing developers to focus on the machine learning models instead of all the surrounding kerfuffle that takes a lot of time but doesn’t directly contribute to the application itself.

Screenshot of in action. Photo credit: slay

“You can connect a data source with the product. This could be your database or an S3 bucket of data that you want to send to a machine learning model. And then the machine learning model – just some Python code – finds predictions in the data. We put that in an API that does things like validate the input the user puts in or do some processing on the output before sending it back to the user,” explains Mernit. “These components form a machine learning application. So if someone was doing these things by hand, they usually had to set up a web server themselves. They would have to set up a versioning system, they would have to set up a way to monitor the model. And all of that means a lot of hard work. We do all this for the user. They just need to focus on where their data is coming from and what kind of model they are using. The rest is done for them. In short, we eliminate all the glue that goes into the machine learning development process.”

The platform sees itself as a GitHub for ML – and makes it easy to fork existing machine learning recipes for use in applications.


About Author

Comments are closed.