I build AI systems that people can rely on

I work where large language models meet real business data. For the last 2.5 years I've designed and shipped enterprise Generative AI platforms for healthcare and retail organisations: multi-agent systems built on LangGraph and the Model Context Protocol, retrieval-augmented assistants, and conversational analytics that let people ask their data questions in plain English. Platforms I've built are used by more than 500 business users.

Most of my effort goes into the parts that make AI dependable rather than impressive for a week: guardrails and PII redaction, evaluation and tracing, latency budgets, and knowing exactly what each agent costs to run. I ship on FastAPI, Docker and Kubernetes across Azure and AWS, and integrate with the data platforms teams already use, such as Snowflake and Databricks.

I now take on independent projects too, helping founders get from idea to a working product, and helping larger teams harden, evaluate and scale the AI systems they already have.