BSc Computer Science graduate from the University of Bristol, building applied systems across VR, machine learning, computer vision, robotics and graphics.
ML/AI Prediction Micro-services Application for a Client, DigitalU3
Oct 2024 – Apr 2025
Developed an ML/AI micro-service layer for real-time detection of fire, smoke and oil leaks in an IoT platform for which DigitalU3 received an Intel innovation award.
Trained and fine-tuned a YOLOv5 CNN on 900+ labelled images for fire and smoke object detection.
Built REST API endpoints using Python and SQL, developed the React frontend, and deployed using AWS and Docker.
Worked with kanban workflows, pull request reviews, and client/mentor feedback cycles.
PythonYOLOv5ReactSQLDockerAWS
03projects2 case studies · 3 notes
// everything else — select a row to expand
name
domain
stack
status
vision
opencv · python
shipped
Classical computer vision pipeline covering object detection, stereo geometry and 3D reconstruction.
Viola-Jones dartboard detector with colour thresholding and Hough circles.
Epipolar lines, correspondence matching and triangulation.
3D sphere position and radius estimation under noisy camera pose.
ml
pytorch · pymc
shipped
Implemented and compared core ML methods across classification, clustering, regression and sequence modelling.
K-Means, GMMs, neural networks and HMMs.
Deterministic and probabilistic approaches, including EM-style methods.
Bayesian modelling explored using PyMC.
graphics
c++ · sdl2 · glm
building
A software rasteriser built from scratch in C++ using SDL2 and GLM, without relying on OpenGL or DirectX for the core renderer.
OBJ/MTL parsing, vertex transforms and perspective projection.
Triangle rasterisation, texture mapping and depth buffering.
Coordinate transforms, clipping and camera projection.