RL Environment Engineer, SpaceXAI
Jan 2026 – Present- Built LLM evaluation and training pipelines for Grok model development using Python, FastAPI, LangChain, LangGraph, Langfuse, pgvector, and Docker covering observability, agent tracing, retrieval augmented workflows, and structured eval loops for model iteration.
- Designed and shipped 20+ RL environments to evaluate LLM agent behavior, each with custom rubrics and ground-truth references, using Python, FastAPI, TypeScript, React, Next.js, PostgreSQL, and Docker enabling reproducible, scalable agent benchmarking.
