Anyscale, which helps apps run efficiently across distributed infrastructure with the Ray open source framework, raises $100M Series C led by a16z and Addition
The UC Berkeley computer science labs, from which seven academics formed $38 billion data startup Databricks, have minted a second billion-dollar business.
Context & Ripple Effects
Anyscale is following the playbook its founders' Berkeley neighbors already proved: take an open source distributed-computing framework out of the UC Berkeley labs and wrap a commercial platform around it. The $1B round that valued Databricks at $28B showed how far that template can run for Apache Spark; Ray is now getting the same treatment.
The raise also extends Andreessen Horowitz's position as the connective investor between the two companies — it led Anyscale's 2019 Series A and backed Databricks through its Series C and beyond, making this third check on Ray a bet on repeating the Spark outcome.
First-order effects
- Anyscale moves from a $40M Series B to a nine-figure war chest just fourteen months later, giving it capital to scale the commercial platform around Ray while the framework's developer base is still forming.
Second-order effects
- Databricks' trajectory — from Berkeley lab project to a company later reported above a $7B revenue run rate — sets the benchmark investors will hold Anyscale to, pressuring it to convert Ray adoption into enterprise contracts on a similar curve.
Third-order effects
- If the Spark-then-Ray pattern holds, UC Berkeley's systems labs become a repeatable origination channel for infrastructure unicorns, with a small set of repeat backers like a16z capturing the sequence early each time.
The trend: Academic open source frameworks are being converted into venture-backed infrastructure platforms, with UC Berkeley labs and repeat lead investors like Andreessen Horowitz as the recurring origin points.