Staff AI Engineer

Drata is the proof layer that helps companies earn and keep the trust of their users, customers, partners, and prospects. As a Staff AI Engineer you'll help shape how intelligent systems power trust-critical enterprise workflows, owning systems end-to-end from early research to production deployment. This is an end-to-end ownership role that shapes how AI systems are built company-wide, influencing technical direction across LLMs, retrieval systems, and agentic workflows.

What you'll do:

• Shape the Architecture: Design and own production AI systems end-to-end (LLM pipelines, RAG, reranking, vector stores, orchestration). Make thoughtful build/don't-build decisions based on real data. Evolve the AI stack over time, from model infrastructure to workflow orchestration to evaluation tooling.
• Raise the Quality Bar: Design evaluation systems that measure retrieval quality, reasoning accuracy, and end-to-end performance. Build tooling that helps the team iterate confidently and catch regressions early.
• Investigate Deeply & Decide with Evidence: Analyze production outputs to identify failure patterns and root causes. Turn complex findings into clear technical recommendations. Determine when to advance versus pause based on quantitative analysis.
• Lead Across Teams: Be the go-to technical voice for AI architecture decisions. Influence standards for how LLM systems are built, tested, deployed, and monitored. Mentor senior engineers through design reviews and hands-on collaboration. Partner with product and compliance teams.
• Build Responsible, Production-Ready AI: Ship systems optimized for latency, cost, reliability, and auditability. Embed safety guardrails, confidence thresholds, and human-in-the-loop workflows. Ensure outputs remain traceable and explainable.

What you'll bring:

• 10+ years of software engineering experience, including 3+ years working directly on ML/AI systems
• Real ownership of production LLM systems
• Deep experience with RAG, embeddings, reranking, vector databases (Pinecone, FAISS, Chroma, etc.), and agentic workflows
• Experience designing evaluation frameworks and using quantitative analysis to improve system performance
• Strong Python skills (TypeScript is a plus)
• A track record of making architectural decisions that shape team direction
• Production AI systems experience: observability, reliability, cost tradeoffs
• Ability to break down ambiguous, high-stakes problems into structured investigations

Nice to have:

• Compliance, security, or regulated domain experience
• Enterprise data platforms or Snowflake-based analytics familiarity
• Orchestration systems experience (Temporal or Airflow)
• LLM evaluation platform experience (e.g., Braintrust)
• Technical community contributions or published work

Remote across the U.S.; hybrid option from the San Francisco office (Tuesday-Thursday). Base salary range $200,700 - $271,500.