Lead LLM Engineer

About the role

We are looking for a Lead LLM Engineer to make our AI outputs reliable, fast, and indispensable in real workflows. You will own the end‑to‑end quality of large language model (LLM) systems, turning vague user problems into structured, measurable AI solutions that improve continuously.

Key responsibilities

• Design and evolve the LLM/agent architecture.
• Own output quality across core use cases such as email generation and document analysis.
• Build evaluation systems, including datasets, metrics, and regression detection.
• Drive rapid iteration loops using production data.
• Improve retrieval, reasoning, and tool usage within the models.
• Ensure production reliability, managing latency, failure modes, and fallback mechanisms.
• Collaborate directly with product teams and founders to define what to build and why.

Required profile

• Experience building LLM‑based systems that run in production, not just demos.
• Deep understanding of Retrieval‑Augmented Generation (RAG), tool integration, agents, and structured outputs.
• Ability to design full pipelines beyond prompt engineering.
• Evaluation‑driven mindset: define quality metrics, create real‑usage datasets, and run continuous evaluations to prevent regressions.
• Strong debugging skills: trace issues across retrieval, prompts, and model behavior.
• Fast iteration capability: move from problem identification to improvement within hours or days using logs, traces, and data.
• Sound judgment on when to use agents versus pipelines and how to balance complexity with reliability.

Required skills

• Python (FastAPI)
• PostgreSQL
• Google Cloud
• LangGraph / LangChain
• PostHog (product analytics)
• Langfuse (LLM tracing)
• Azure OpenAI (LLM APIs)