I've built an MCP client-server system from scratch and a 12-node LangGraph multi-agent platform over more than 1 million PostgreSQL rows, with ChromaDB retrieval and answer verification against ground truth.
At Express Housekeeper, I traced a four-month compliance KPI decline to its root cause and restored 100% audit compliance. I also built SQL and Python pipelines and reusable validation scripts that reduced manual reporting effort by 30–40%.
My work combines Python, SQL, classical machine learning, LLMs, RAG, FastAPI, and MLflow. I hold an M.Sc. in Data Science and Analytics and have contributed NLP and LLM research at Maynooth University's ADAPT Centre.
