You will build the infrastructure that runs agentic development in the cloud: ephemeral, reproducible, isolated environments where many agents work concurrently on real Motive codebases, run real tests against real dependencies, and open real pull requests - without an engineer babysitting a terminal, and without compromising security or blowing up spend. Evaluate, select, and roll out AI developer tools (coding agents, code review, test generation, incident investigation, documentation) and build the internal scaffolding - context, MCP servers, evals, guardrails - that makes them effective on Motive''s actual codebases rather than on a demo repo.