Description

I add AI to software that businesses already use, and I build small AI tools and assistants from scratch. That covers things like letting an AI answer questions from your own documents, automating a repetitive task that currently takes up someone's day, automating emails, pulling data out of PDFs, connecting AI to the systems you already run, and checking the results are correct rather than just plausible.

I work mainly in Python and Rust, and in TypeScript and React Native when a project needs a website or a mobile app. Alongside client work I build and release my own software, including open-source developer tools and my own iOS, Android and web apps, so I'm comfortable taking something from a first idea through to a finished product people use day to day.

Before this I spent two years at Diamond Light Source, the UK's national synchrotron, writing machine-learning and high-performance computing software for X-ray imaging, on top of a physics degree and a PhD. That background shapes how I work: I test what I build, and I don't hand over code I can't explain.

Typical projects: adding an AI feature to an existing product, building an automation around a repetitive task, setting up a system that answers questions from your documents, or making sure an AI tool gives accurate answers before it goes live. I keep projects clearly defined, agree a fixed price where I can, and work in days and weeks rather than months.

Industry field of expertise

Languages

English
Native or bilingual

Workplace preferences

Can work on-site

Munich (up to 10km)

Independent
AI / Software Engineer
DIGITAL AND IT
November 2025 - Today (7 months)
Munich, Germany
Build AI tools and automations for clients, and release my own open-source software.
Recent open-source work: a fast local code-search engine and a clipboard-history tool for AI assistants, both written in Rust and published as packages, plus a benchmark tool for testing large language models.
Built and released my own iOS, Android and web apps, covering everything from the AI features through to the interface.
Work across Python, Rust and TypeScript, with React Native or Swift for mobile.
Python Rust LLM AI Agent Typescript
Diamond Light Source
Postdoctoral Research Associate
RESEARCH
October 2023 - November 2025 (2 years and 1 month)
Oxford, UK
Built GPU-accelerated software in Python and JAX to process and reconstruct large 3D X-ray imaging data, and turned it into tools that visiting scientists could run themselves.
Developed machine-learning and computer-vision methods for that data, including automatic image segmentation and reconstruction, on datasets running into the terabytes.
Set up automated, repeatable processing on a high-performance computing cluster, with versioned settings and batch jobs, so results could be reproduced and handed over to other people.
Supported live experiments with academic and industrial teams, diagnosing data problems as they happened and adjusting the processing when conditions changed.
Wrote the technical documentation and step-by-step procedures, and presented progress to management and project partners, including a collaboration with the University of Oxford.
Python Machine learning Computer Vision JAX GPU Computing
University of Manchester
PhD Researcher, Materials Science
RESEARCH
October 2018 - September 2023 (4 years and 11 months)
Manchester, United Kingdom
Built a machine-learning system that automatically labelled features in 3D X-ray scans, cutting a task that took hours of manual work down to seconds per scan.
Used a transformer-based model (SegFormer) and compared it against standard approaches, packaging it so the training could be rerun on a GPU.
Wrote custom Python tools to analyse large, multi-part imaging datasets.
Ran long experiments at national research facilities, managed a £30,000 budget, and won access to specialist equipment through written proposals.
Reported results to industrial sponsors and mentored newer students.
Python Pytorch Machine learning Computer Vision Data analysis