- Help decide what gets built, why it matters, and how it works end-to-end.
- Work on systems that are already live across thousands of developer hubs and define what they become next.
- Design end-to-end AI systems combining LLMs, embeddings, retrieval pipelines, and structured outputs into reliable product features.
- Build grounding and retrieval layers: indexing docs, OpenAPI specs, and customer data to support accurate, context-aware generation.
- Define evaluation systems: measuring accuracy, latency, cost, and user impact; building feedback loops to continuously improve quality.
- Handle real-world complexity: malformed specs, inconsistent docs, edge-case APIs, and ambiguous user queries.
- Make architectural tradeoffs: latency vs. quality, cost vs. coverage, deterministic vs. generative approaches.
- Own production reliability: observability, fallbacks, rate limiting, and safe degradation when systems fail.
- Work across the stack: from model orchestration and backend systems to the UI surfaces that expose them.
PythonJavascriptTypeScript+1 more