MSc Artificial Intelligence and Adaptive Systems at the University of Sussex. I build mobile test automation that survives contact with real devices, and I like algorithms close enough to the metal to read in one sitting.
| Project | What it does |
|---|---|
| maestro-plus | MCP server layered on Maestro's official one: multi-device pools, assertions on run outcomes, and failure diagnosis that separates application defects from test defects. Built on MCP Python SDK v2. CI green on 3.10/3.12. |
| cv-from-scratch | Classical computer vision algorithms from scratch in pure NumPy — Canny, Hough, morphology, Harris — each paired with the what/why/how notes textbooks skip. 65 tests, sub-second suite. |
- MSc research: biologically-inspired convolution front-ends and visual robustness (code repository in progress)
- Maintaining both projects above — issues and real failure samples are welcome, especially ones with a known root cause
Python · pytest/ruff · GitHub Actions · Maestro/ADB · Docker · PyTorch · Flask