Software Developer · Backend / Full-Stack
Predicted First Class BSc Computer Science student building full-stack, backend and AI-assisted software across .NET, Angular, React, SQL and Python.
Final-year Computer Science student at Queen Mary University of London, on track for a Predicted First Class BSc in June 2026 — focused on backend, full-stack and AI-assisted software.
Designed and implemented a modular deep learning pipeline that takes a single meal image and produces item-level food predictions, segmentation masks, confidence scores, and calorie-per-gram nutrition lookup outputs.
Manual dietary logging is time-consuming and error-prone — users have to remember every food item and estimate portions themselves. Real meal images make automation hard too, since foods overlap, look visually similar, and often contain multiple ingredients.
Built a modular image-assisted dietary assessment pipeline. RF-DETR was fine-tuned for one-class food detection, SAM2 generated instance masks from the detected regions, and ViT-B/16 classified the segmented food crops. Predicted labels were then matched against a nutrition lookup database to return calorie-per-gram outputs.
An Azure-backed legal contract analysis tool built during my Willis Towers Watson internship. Owned the database schema, MSAL Single Sign-On with Azure AD, SHA-256 file validation and tested C# claim-year modules.
A property listings and reviews application built with the MERN stack. Co-designed the UI/UX in Figma, paired-programmed weekly via scrum, and presented the final solution to FDM consultants.
Designed and built a clean, functional homepage for Lucy A Raymond & Sons holdings, delivered as a polished client-facing presentation that enhanced the company’s professional image.
A Java Swing GUI simulator with horse confidence/breed mechanics and a fully functional betting system. Architected using SOLID principles and object-oriented design.
Programmed a robot to follow a white line on an unmapped path using a light detector, micro:bit and JavaScript — built collaboratively at Imperial College London summer school.
Led a project using React and Tailwind to build a functional homepage website for the holdings company, delivered as a polished client-facing presentation. Configured 4 employee laptops by assessing individual requirements for a smooth setup and improved productivity.
Evaluated insurance risk applications by analysing applicant information and determining risk levels. Co-ordinated senior underwriters on 20 complex cases including 3 Latin American countries, prepared detailed risk reports, and achieved a 5% increase in client response time by applying technological tools.
Guided students through debugging methods in Java, gave clear instructions on programming tasks, and emphasised good coding practices including reliability, commenting and concise documentation.
Attended a week-long University Mathematics summer school exploring advanced ideas around infinity — countable vs. uncountable sets, why the infinity of reals exceeds that of the integers, and Hilbert's Hotel-style problems. Also joined a J.P. Morgan session on CV improvement, interview preparation, and early-career development.
Attended a JLR Formula E experience in Oxford, learning about the company's values, engineering culture, and working life. Also joined a Google event exploring Gemma, speaking with software engineers and site reliability engineers about their day-to-day responsibilities, technical areas, and career paths.
Completed a J.P. Morgan mentorship focused on software engineering careers in financial services — covering different areas of J.P. Morgan markets, the engineering skills used in banking, and key financial concepts such as hedging strategies.
Explore my projects, GitHub, and technical work.