I build and run AI systems in production. Most of my work is around LLM apps, generative pipelines, and agent workflows, along with the backend pieces that keep them stable — retrieval, data pipelines, and evaluation. I mainly work in Python with PyTorch and Hugging Face. I build APIs with FastAPI or gRPC, and deploy using Docker, Kubernetes, and GPU stacks like vLLM and Triton. I also use tools like LangChain or LlamaIndex. I’m used to owning things end-to-end, from design through to production. If you’re working on something real in this space, happy to connect.