I build software for evaluating AI honestly.

I’m an MSc Artificial Intelligence candidate at the University of Sheffield, with a BSc in Computer Science from Queen’s University Belfast. Most of my recent work is data and ML engineering with an evaluation backbone: a 41.4M-journey PySpark → dbt pipeline quantifying London transport disruption, a Dagster-orchestrated open-data engine behind a public API and website, an LP/MILP energy-scheduling system with a forecast-vs-actual retro, end-to-end PHM MLOps on the NASA and ESA benchmarks, and release-gates and red-team harnesses for AI agents.

I build systems whose decisions you can audit: tested pipelines, honest evaluation, and evidence you can trace back to where it came from. Most of my projects combine Python and SQL, dimensional models with data tests at every layer, structured evaluation, and typed APIs with simple interfaces that both technical and non-technical readers can use. I try to understand the failure modes as carefully as the wins, and to check whether a number is well-posed before trusting it.

I’m looking for graduate or entry-level roles in data engineering, ML engineering, analytics engineering, and AI evaluation or safety engineering — work where pipelines and models people can trust are what matter. I’m about as comfortable on a Spark-scale backfill as on a dbt warehouse, an ML forecast, or an evaluation harness.

I’m available from October 2026.

The fastest way to reach me is by email, or through LinkedIn.