Founded in 2024, Blacksmith set out to make continuous integration—the automated building, testing and validation of software changes—less arduous. Its premise was straightforward: code can move only as fast as teams can verify it is safe to merge. CEO Aditya Jayaprakash said the company was built to make CI “less painful.”¹

That problem has accelerated with the arrival of coding agents. Jayaprakash argues that tools such as Cursor, Codex and Claude Code have made writing software “extremely easy,” shifting the pressure point to a harder question: “how do I actually validate all this code?”² In Blacksmith’s framing, the AI boom is creating a testing bottleneck, not eliminating the need for developers’ guardrails.

On Wednesday, the company announced a $66 million Series C led by Sequoia, valuing Blacksmith at $1 billion. The leap is striking: just two months earlier, a $45 million Series B put its value at $550 million; a year ago, after its Series A, it was valued at $60 million.¹ Total funding now stands at roughly $125 million.

The operating figures explain the investor appetite. Blacksmith says it has run more than 90 million CI jobs and grown annual recurring revenue 17-fold. Its customer base has surpassed 10,000 companies, while weekly validation checks rose from 5 million at the start of the year to 90 million.²

Blacksmith’s answer is capital-intensive: rather than rely entirely on cloud providers, it buys and operates tens of thousands of servers in data centers worldwide. Much of the new funding will go toward more compute, Jayaprakash said.² That gives the startup performance control, but leaves it exposed to shortages of compute and memory—and to rivals. GitHub already sells hosted CI machines, while AI coding vendors are increasingly building testing into their agents.

Jayaprakash’s counterargument is blunt: stronger models generate more code, and more code means more checks. “As agents write more of the world’s software, the question that matters is whether it can be trusted.”¹