I will integrate an llm or ai feature into your app

Most "AI integrations" are a demo that breaks in production. I ship LLM features that survive real traffic: guardrails, evals, cost control, retries, and monitoring, the parts that actually matter.

I am a Staff/Principal engineer with 12+ years shipping production software. I led an 8-person platform team running over $5B/year in AI lending decisions, where I built the eval harnesses and model-promotion gates that let us ship LLM and ML changes safely. Today I ship LLM products solo on a live stack (Fly, Vercel, Neon).

What I do:

• Integrate OpenAI, Anthropic Claude, or open models into your app with clean Python or TypeScript
• - Proper prompt design, structured outputs, streaming, and error handling
• - Retrieval and tool use where you need it, plus cost and latency controls
• - Evals and monitoring so you know when output drifts, not just when it crashes
• - Production code and a clean handoff, not a notebook

This is hands-on build plus consultation. Tell me your stack and what the feature should do, and I will scope it precisely.

No unlimited revisions, which keeps scope clean. Message me before ordering so we can scope it right.