Currently building AI systems β multi-agent pipelines, RAG architectures, and LLM backends. When I'm not debugging LangGraph state graphs, I'm probably vibe-coding something at 2am.
- Agentic AI β stateful multi-agent workflows with LangGraph. basically FSMs but make it LLMs
- RAG β hybrid retrieval, reranking, chunking strategies that don't make the model hallucinate
- Evaluation β LangSmith tracing and evals to build deterministic testing pipelines at scale
- MCP β Model Context Protocol for distributed tool integration
- Prompt Engineering β crafting precise system instructions, iterative context optimization, and testing edge cases
open to GenAI roles. if you're building something that needs an LLM backend, let's talk.
(personal website: under progress with vibe coding)
