Dr Stephen J Newhouse

Applied AI and data science. Bioinformatics, knowledge graphs, agentic AI.

Hi, I’m Stephen.

I’ve spent 20+ years in bioinformatics, genomics, and life-sciences data science. That sounds like a long time until you remember how much the field has changed in just the last five. The questions stay the same though: how do we make sense of complex biomedical data, and how do we turn it into something that actually helps people?

For five years I led Professional Services at Biorelate, a UK biomedical AI company. In practice that meant making sure the science connected to what clients actually needed: accelerating a drug-discovery pipeline, untangling a messy real-world data workflow, or helping a team understand what they’d got and what they could do with it. Since June 2026 I’ve been working as an independent consultant and adviser.

Before this I worked across academia, the NHS, and pharma, building and leading teams, shipping ML models, and trying to make large-scale data projects land well in the real world.

Somewhere along the way I appear to have become an Agentic Engineer. Not by design, more by osmosis and a stubborn refusal to fall behind the LLM Joneses. Right now I’m particularly interested in where knowledge graphs, RAG, and multi-omics data are heading, and what that means for personalised medicine.

Always happy to connect with people working at the intersection of data, biology, and drug discovery. → LinkedIn · Email

Recent outcomes

  • Sole engineer on a production agent SDK and MCP server over a biomedical knowledge graph (memory, tool-use, evaluation, knowledge-graph reasoning), integrated by a top-10 global pharma into one of their internal agentic AI platforms.
  • Sole architect of a multi-agent target safety assessment system: knowledge graph + LLM evidence retrieval, programmatic citation, citation-traced reports in hours rather than weeks.
  • Built and ran a Professional Services and consulting function from zero.

What I’m working on