‘Reinforcement Learning Gym’ Startup, Buoyed by Labs’ Appetite For Training Data, Reaches $750 Million Valuation

AI developers are getting more creative in how they acquire data to train AI models.

For instance, they’re paying startups to develop copies of popular apps, like Salesforce or Excel, to teach models how to use the apps—otherwise known as https://www.theinformation.com/articles/anthropic-openai-developing-ai-co-workers">reinforcement learning environments or gyms.

One beneficiary of this trend is Fleet, a two-year-old startup that develops and sells those RL environments.

The startup is now in talks to raise at least $50 million in new funding at a valuation of around $750 million, including the investment, according to a person with knowledge of the talks.

That’s more than a seven-times bump from its seed round valuation, which was less than $100 million, the person said.

Bain Capital Ventures is in talks to lead the latest round, while existing investors Sequoia Capital, Menlo Ventures and SV Angel plan to participate, the person said. (Spokespeople for Sequoia and Menlo declined to comment, while Bain and SV Angel didn’t respond to a request for comment.)

One driver behind the latest round is the company’s strong revenue growth.

Fleet has grown from $1 million in annualized revenue as of late last year to more than $60 million in recent weeks, the person said. (The company calculates annualized revenue by multiplying the last quarter’s revenue by four.)