Undertook multiple machine learning computer vision projects from initial data aggregation and modelling to evaluation and a final write-up of the project. This includes: Project 1) Instance segmentation project to detect damaged parts on a car. Training done using the Mask2Former model. Part of a PoC project, targeting insurance companies to help with insurance claim cases.Project 2) Object Detection project to identify headlights/taillights of a car, as part of a quality assurance project for an automotive client. Project involved utilising YOLOx, achieving a mAP score of 66.2 which compares to the standard benchmark score of 40.5.Project 3) Object detection project to identify different plant diseases to help combat common issues faced within the Agriculture Industry. The model trained is the Faster-RCNN model.Project 4) Created a multi character chatbot using LLMs, and deployed into a dashboard using Streamlit. This was used for product research purposes, with the aim to simulate end consumer.Project 5) Conducted research and collaborated on a reinforcement learning trading bot project, which utilized PPO algorithm with CNN policy.

Main Project: Created a Natural Language Processing chatbot to help internal analysts access bond information more efficiently.Created a budgeting website for NatWest customers, with full front and back-end capabilities on a live server.Raised £500 for the Prince’s trust by doing a group 6-legged walk around London.