AI / ML Engineer
Principal-level ML engineer working on LLM systems, applied machine learning in production, and self-hosted AI infrastructure. I turn research-grade ideas into reliable, maintainable software.
Available for consulting and contract work
01 / About
I am an ML engineer with a strong data-science foundation, now focused on the engineering side of AI: turning models into production systems that hold up under real load and real users.
My recent work centers on large language model systems, retrieval-augmented applications, fine-tuning, and the infrastructure that runs them. I keep a self-hosted GPU cluster at home for hands-on experimentation, so the things I recommend are things I have actually run end to end. I care about clean interfaces, reproducibility, and code other people can maintain.
02 / Skills
03 / Projects
A few things I have built. Each is something I designed, wrote, and ran.
A home GPU cluster serving open-weight models over an OpenAI-compatible API. vLLM for throughput, a thin FastAPI gateway for auth and routing, and Proxmox for isolation. Used daily for experimentation and fine-tuning.
A RAG pipeline over a private document set: chunking, embeddings, a vector store, and a reranking step that measurably cut hallucinations. Built to run fully offline on local hardware.
An end-to-end pipeline taking a model from notebook to a containerized service with monitoring, batch and real-time scoring, and reproducible retraining.
04 / Services
Design and build RAG pipelines, agents, and LLM-backed features. From prototype to a service you can actually run and trust.
Take a model out of the notebook: packaging, serving, monitoring, retraining, and the glue code that keeps it reliable.
Self-hosting open-weight models, GPU and inference setup, and pragmatic build-vs-buy advice for teams getting started with AI.
05 / Contact
Open to consulting, contract work, and interesting problems. The fastest way to reach me is email.
06 / Latest Blog Posts