Factory AI Raises $150M at $1.5B Valuation to Build Autonomous 'Droids' That Replace Dev Teams

Factory AI has closed a $150 million Series C at a $1.5 billion valuation, with Khosla Ventures and Sequoia leading the round. The San Francisco company, founded in 2023, builds autonomous AI agents it calls “Droids” — software that handles the full development lifecycle end-to-end: writing code, running tests, reviewing pull requests, and pushing deployments.

The pitch is blunter than most AI coding tools: this isn’t about autocomplete. Droids are designed to own tasks entirely, not assist humans who still do the work. Factory positions the product as a replacement layer for repetitive engineering work, not a copilot sitting next to a developer.

Who’s already using it

The customer list is notable. NVIDIA, Adobe, and MongoDB are production users. Factory claims hundreds of thousands of developers touch its platform in some form, though the enterprise contracts with those three names carry most of the commercial weight. The company hasn’t disclosed ARR figures, but the valuation jump from its previous round implies strong retention data.

Droids operate via a multi-agent architecture. One agent plans the task, another writes the code, a third runs the test suite, and a fourth handles the pull request and deployment pipeline. Each agent can call tools, read repositories, access CI/CD systems, and escalate to a human when confidence drops below a threshold. The system is designed to be auditable — every decision is logged with reasoning.

The broader context

Factory is raising into a crowded category. GitHub Copilot Workspace, Cursor’s background agents, Cognition’s Devin, and dozens of smaller players all claim some version of “autonomous coding.” What distinguishes Factory’s approach, at least in its marketing, is an explicit focus on the full lifecycle rather than single-file edits or greenfield projects.

The $150M round brings Factory’s total raised to roughly $200M. Sequoia’s involvement — the firm also backed Cursor and GitHub — signals the category hasn’t hit saturation from a venture perspective. The memory of Devin’s rocky public launch in 2024, where real-world benchmarks fell far below demo performance, still hangs over the space. Factory has largely avoided those public comparisons by staying enterprise-focused and releasing quietly.

What this means for engineering teams

The honest answer: it depends on what your team’s engineering bottleneck actually is. Droids perform well on tasks with clear specifications and robust test coverage. They struggle on ambiguous requirements, large existing codebases with poor documentation, and anything requiring business context that isn’t written down.

For companies willing to invest in the integration work — connecting Factory to their CI/CD pipeline, adding test coverage to match — the ROI case is real. Oracle’s 30,000 headcount reduction, announced this week for AI-related restructuring, is an extreme version of the pressure the entire industry is feeling. Factory is betting the middle path — not replacing all engineers but replacing 60–80% of the ticket queue — is where enterprise money actually goes.

The $1.5 billion number represents a bet that this middle path scales. The Khosla and Sequoia check sizes suggest two of the more disciplined venture firms in Silicon Valley agree.