Zeke Burningham is an aeronautical engineer and the founder of Markethalo, combining rigorous modelling, software development and applied AI to solve complex technical problems.
I'm Zeke Burningham — an aeronautical engineer (MSc, Imperial College London; First Class BEng, Exeter) working on data-heavy, physics-constrained problems where the answer has to actually work. My MSc research models the climate impact of a next-generation blended-wing-body aircraft using reanalysis meteorology and contrail physics — the same rigor now goes into AI and automation systems built for technical industries.
The difference is domain fluency: models grounded in real engineering constraints, pipelines built to be maintained, and prototypes that de-risk a decision instead of decorating a pitch.
Internal tools that automate modeling, analysis, and reporting workflows — cutting hours of manual work out of the technical loop.
Turn manual, spreadsheet-bound processes into reliable automated systems with validation, logging, and clean handoffs.
Physics-aware models and forecasting built for real operational decisions — grounded in constraints, not just correlations.
Proof-of-concept builds that de-risk a bet before you commit budget — a working system you can test, not a pitch.
Fuel-burn-only CO₂ comparisons miss non-CO₂ effects — like contrail cirrus — that can dominate an aircraft's total climate impact.
Built a modelling pipeline on the pycontrails CoCiP framework driven by ERA5 model-level meteorology, deriving a modified Schmidt-Appleman contrail formation criterion for the aircraft's boundary-layer-ingesting propulsion.
A combined CO₂ + contrail + NOₓ effective radiative forcing comparison against conventional narrowbody baselines — Imperial College London, 2026.
A 14-state VTOL hover model needed a controller that stayed robust across the full flight envelope, not just at one operating point.
Designed and validated H-infinity loop-shaping controllers, with full frequency-domain analysis and stability margins produced to aerospace review standard.
52% reduction in worst-case gain across the flight envelope — University of Exeter, First Class Honours.
Wanted to test an ensemble quant strategy against live markets end-to-end, without manual execution.
Built a full-stack paper-trading platform — React/Vite frontend, Express backend, real-time Finnhub WebSocket feed, and an ensemble model combining multiple signal sources.
A working system processing live data and executing paper trades end-to-end — self-initiated and self-built.
Map the real problem and where the time actually goes. Scope tightly.
Ship a working system in short iterations, testable from week one.
Get it into production cleanly, with the docs and handoff to run it.
Iterate on real usage — tune, extend, and keep it reliable.
Tell me the problem in a sentence or two. If it's a fit, I'll come back with a scoped approach — not a sales call.