~/portfolio — building in AI, summer 2026
MSc Advanced Computer Science @ University of Manchester
Building AI projects, running experiments, and figuring out what actually works.
01—ABOUT
I'm a recent First-Class Computer Science graduate from Newcastle University, starting an MSc in Advanced Computer Science at the University of Manchester. I'm interested in building practical AI applications, exploring new ideas, and learning by creating projects from scratch. This site is where I document that journey.
LOCManchester, UK
02—PROJECTS
INDEX
01/ 00
SHIPPED
Warns ICU nurses hours before a patient deteriorates, scoring five risks every hour and beating the NHS's NEWS2 score at the same alarm rate. Runs entirely on IBM LinuxONE, so patient data never leaves the machine. Built for the IBM Z Datathon 2026.
Time-travel debugging for AI agents: replay, inspect, fork and compare every decision an LLM agent makes, from recorded traces that replay exactly the same way every time.
Three AI personas, including a screen-reader user, browse a live site in parallel and return a ranked list of usability issues, each tied to the exact screenshot and reaction that caused it.
My First-Class dissertation: does grounding an LLM in a researcher's published papers, or pairing two researchers from distant fields, make it come up with better and more novel research ideas?
Snap any object with your phone and it wakes up as a cheeky Mancunian character that rap-battles you, roasts your comebacks and remembers you next time. Built in a day at Hacktoberfest Hack Day Manchester.
Tap out a rhythm on your keyboard and a local AI producer turns it into a full track with drums, bass, synths, a title and vocal ideas. No music knowledge needed. Built in a day at the Bradford Agentic AI Hackathon.
03—CONTACT
Open to internship and graduate opportunities in AI/ML, and happy to hear from anyone who wants to compare notes on a project.
