| 01 |
Oxford Physical AI Hackathon Oxford AI Society · Track 2 · SO-101 + Amazing Hands · Neuracore Stack Only team to assemble the Amazing Hand hardware end-to-end. Shipped three live demos in 24 hours — real-time hand mirroring via imitation learning, sign language recognition, and a Meta Quest VR control interface. |
| 02 |
Patient Continuity in Crisis Zones Imperial College London · Crisis Zones Hackathon Built a centralised hospital system for conflict regions with biometric patient matching via face-to-vector embeddings — no raw image storage. Offline-first architecture with local caching for degraded and unstable networks. |
| 03 |
AssertionOS Encode · SpoonOS Hackathon AI-powered multi-chain wallet security agent with real-time threat detection for DeFi. Built the on-chain reasoning module and tool-use integration layer. |
Adaptive Race Intelligence System — a physics-grounded, ML-augmented F1 strategy engine that evaluates pit decisions per tick using Monte Carlo confidence intervals and narrates the optimal call with a quantified lap-time delta.
PIPELINE ───────────────────────────────────────────────────────────────────
FastF1 Ingest → PostgreSQL → Bicycle Model + Tyre Degradation + Fuel Burn
↓
XGBoost ML Residual
(race-by-race CV)
↓
Monte Carlo Action-Search Engine
↙ ↘
Strategy Output LLM Narration
delta · CI · pit window radio-call format
↘ ↙
Streamlit Dashboard
─────────────────────────────────────────────────────────────────────────────
| Phase | Scope | Status |
|---|---|---|
| 1 · Foundations | FastF1 pipeline · PostgreSQL · Streamlit | ✅ |
| 2 · Predictor | Bicycle model · Tyre degradation · XGBoost residual | 🔄 |
| 3 · Counterfactuals | Perturbation engine · Tiered tick architecture | ⏳ |
| 4 · Narration | Local LLM · Conformal calibration · Strategy backtest | ⏳ |
→ Execution Plan · Build Log · Repository
Jan 2026 – Mar 2026 Student Leader · Electech Lab
Taught robotics, logic gates, and coding to primary-school
students at St Andrews C of E using Tinkercad + Code Monkey.
Jul 2024 – Aug 2024 PHP Intern · Tally Marks Consulting
Optimised database operations and wrote complex SQL queries
for enterprise data management systems.
status = {
"degree" : "BSc Computer Science · Brunel University London",
"building" : "ARIS — F1 race strategy AI (physics + ML + LLM narration)",
"studying" : ["ML theory", "vehicle dynamics", "conformal prediction"],
"exploring" : ["VLA models", "imitation learning", "physical AI systems"],
"open_to" : ["placements", "internships", "AI/ML · robotics · software engineering"]
}

