Founding AI Engineer at $3.5M-funded SF AI startup

As the first engineering hire, you will build a policy-driven runtime that executes high-stakes B2B commercial workflows. Unlike simple LLM wrappers, you'll architect a deterministic execution layer where models negotiate real contracts and protect revenue. Reporting to a veteran CTO, you'll own the agent pipeline, state management, and safety guardrails.

• Design and evolve a multi-stage agent pipeline covering intent classification, context assembly, reasoning, and response generation as a cohesive system.
• Build the event-driven runtime that governs agent lifecycles, state management, and policy enforcement to ensure safety by construction.
• Develop a robust evaluation framework to define correctness and catch regressions before they reach enterprise customers in high-stakes environments.

• Proven experience building production-grade LLM systems where reliability was non-negotiable and simple prompting was insufficient for the task.
• Strong systems engineering background with proficiency in typed languages like TypeScript, Go, or Rust and an ability to think in complex types.
• Deep understanding of the boundary between model flexibility and deterministic enforcement, with experience in retrieval, context management, and multi-turn reasoning.

• Significant ownership as the first engineering hire with 1% - 1.5% equity and direct access to founders who have scaled multiple successful B2B companies.
• Work on a mission-critical "decision-making" AI system that governs 80/20 commercial outcomes rather than just generating text for chat interfaces.
• Backed by $3.5M in funding from previous investors and led by a CTO whose prior systems verify 80% of US prescription drug returns.