SlipstreamJobsFresh Startup & VC-Backed Jobs

AI Engineer - Assistant Experience

Build - New York, NY, United States - In-office

Apply on the company site

SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.

Build is an AI company creating agentic AI systems for institutional real estate and the built world. The company automates complex development and acquisitions workflows, partnering with major alternative asset investors, developers, and governments. Build is backed by c-suite executives from Blackstone and OpenAI, alongside top venture capital firms. In this role, you will own and expand the capabilities of Dougie, Build's agent harness that executes workflows, grades outputs, and maintains institutional context across sessions. Currently, Dougie operates effectively in controlled settings but lacks the ability to compound learning—what it learns in one run does not reliably improve subsequent runs. Your core accountability is closing this loop. You will wire grading and evaluation directly into the execution path so that runs improve themselves autonomously. You'll build the memory layer that carries institutional context across sessions, enabling Dougie to become smarter with every execution. A key metric you'll own is reducing human-in-the-loop time every week. Key responsibilities include: - Owning Dougie's capabilities end-to-end across Build's ontology for digital labor - Integrating grading and evaluation into the execution path for self-improving runs - Building memory and context retrieval systems that persist learning across sessions - Instrumenting and deliberately reducing human-in-the-loop intervention time - Improving agent orchestration, including planning, tool use, and recovery from partial failures - Creating feedback loops from expert corrections back into system behavior - Debugging non-deterministic systems rigorously with engineering discipline - Owning Dougie's production behavior, including incident response and failure modes The ideal candidate has shipped agentic systems that run unattended with real consequences, and treats memory, context, and retrieval as engineering problems rather than marketing buzzwords. You thrive in high-velocity environments, obsess over excellence, and take full ownership of outcomes. Build values velocity, excellence, long-term thinking, ownership, customer passion, integrity, trust, and measurable results.

Similar roles