Blacksmith Reaches $1 Billion Two Months After a $550 Million Round as AI-Generated Code Overwhelms Continuous Integration

Blacksmith, the platform that runs the build-and-test side of software development, announced a $66 million Series C on Wednesday led by Sequoia at a $1 billion valuation — its third price tag in roughly a year. The company raised a $10 million Series A at $60 million twelve months ago, then a $45 million Series B at $550 million just two months ago, bringing total funding to about $125 million.

Co-founder and CEO Aditya Jayaprakash described the original mission as making continuous integration less painful: the build, test, merge and validate loop that every code change has to survive before it ships. In a blog post announcing the round, he argued that most organizations can no longer keep pace with the volume of CI they need to run — because an enormous share of the code entering those pipelines is now written by AI.

That reframes what the product is for. As autonomous agents generate more software, the bottleneck shifts from writing code to trusting it, and CI is where that trust gets tested. Jayaprakash put it directly: the critical question is whether AI-generated code can be trusted, and the pipeline that answers it has quietly become critical infrastructure.

The company-disclosed numbers backing the round: annual recurring revenue has expanded 17x, and the platform has processed more than 90 million CI jobs — individual automated checks on code changes. Customers named include fintech startup Mercury and AI notepad maker Granola. These are company figures and have not been independently audited, but the funding cadence itself — a 16x valuation step in twelve months — is the clearest market signal.

Blacksmith is riding the same wave as the coding agents it validates. Every leap in AI code generation multiplies the number of builds, test runs and merge checks teams must execute, and AI-heavy workflows are notoriously spiky, hammering infrastructure in bursts rather than steady streams. Tools that absorb that load — and that speed the loop between an agent writing code and a human merging it — have become one of the most contested pockets of developer infrastructure.

What the round really prices is the validation bottleneck rather than generation itself. The labs and startups shipping coding agents get the headlines, but the money is now flowing to whoever owns the moment an AI-written change either earns trust or gets rejected. If agents do become the majority authors of production software, the CI layer sits on every one of those decisions — which is why a $60 million company became a $1 billion one in a single year.