I work on edge AI and the systems that tie devices, data and software together.
A fine-tuned 0.6B language model running traffic signals on a 6 GB consumer GPU through Microsoft Foundry Local, with a safety shield and a signed, hash-chained audit log that maps the controller's actions to UK road-traffic rules after a simulated collision. In simulation it matched MaxPressure, a classic adaptive algorithm, and a fine-tuned model about six times its size. UCL MSc dissertation with Microsoft industry supervision.
-232F3E?style=for-the-badge&logo=amazonaws&logoColor=white){kind=icon}
- Smart-building client onboarding on AWS IoT.
- Dissertation featured above.
- AEM Application Developer @ IBM.
- Shipped RTL-enablement and Carbon Design System v2 migrations for IBM.com.
- Built developer-automation tools with Node.js and Python that improved team efficiency by >70%.
- 💪 Find fulfillment in fitness, 🎸 playing guitar, and
✈️ exploring new cultures through travel. - 🏋️♂️ Recently taken on challenges like 🏃♂️ marathon training, 🚴♂️ cycling long distances, and ⛰️ adventurous treks.
- 📚 Delve into philosophy and 🧠 psychology during quieter moments, seeking profound insights.
- Conversations and collaborations on edge AI, IoT systems and smart infrastructure.
