Software Engineer
BEng Computing @ Imperial College London
I build things that actually ship. AI pipelines, geospatial platforms, full-stack apps. If a problem is interesting enough, I'll build the solution myself before anyone asks.
I'm a software engineer who actually ships. AI-powered medical platforms, urban planning tools, search and rescue systems. Real products, not just demos.
Studying BEng Computing at Imperial College London, I've also done real internships at startups in Hong Kong and won national competitions. I move fast, I own problems end to end, and I don't need hand-holding to get things done.
The thing that actually sets me apart: I build projects for fun. Not to pad a CV, just because I get excited about a problem and want to solve it. That habit shows up in everything I do at work.
Real code under real pressure. Internships, hackathons, side projects. I know what it takes to get something out the door.
Frontend, backend, infra. I'm comfortable going wherever the problem takes me.
I spin up my own projects all the time, exploring new tech and figuring out what's actually possible.
Former Head Boy. Led intern teams. I know how to work with people and keep things moving.
An AI platform that reads and interprets messy PDF medical documents. Built a multi-agent pipeline in FastAPI with async task handling so it could process, summarise, and structure unstructured data at scale, with proper output validation so nothing slips through.
Search and rescue system for lost or injured hikers. Drones with thermal imaging, low-cost stationary sensors on trails, and predictive AI that narrows down search zones before responders even set off. Cuts response time from up to 90 minutes down to 5-15 minutes, at around $2k setup per 20km of trail.
Built a multimodal sensing pipeline across audio, motion, thermal, and GPS data, an AI model that scores risk hotspots, and a route optimisation system that feeds responders live guidance on mobile.
Real-time urban planning tool built at HackLondon. Pulled together multiple external APIs, vector tiles, and constraint datasets into a single parcel-level data model. On top of that, built a RAG architecture with vector search and streaming responses so city planners can actually query the data in plain English.
A Kotlin app that takes geospatial population data and uses it to model simplified transport networks. Runs Dijkstra's and A* to find optimal paths between nodes and spits out journey metrics. Also generates an interactive map so you can actually see what the network looks like.
A news aggregator with a built-in recommendation engine. Built it from scratch because I wanted to understand how personalisation actually works at the system level, not just use someone else's API for it.
Won 1st place out of 700+ students at the Reply Student Clash for an AI platform that interprets medical documents.
Took 2nd place with CityZenith, a real-time urban planning intelligence platform.
Imperial College Alumni Association of Hong Kong Scholarship. 3 recipients from around 60 students.
Given to students who really excelled in their IB Diploma. Proud of this one.
Always up for internships, collabs, or just an interesting conversation. If you're working on something ambitious, reach out.