Scot-Survivor's Portfolio
I have experience across software dev, data science and a little DevOps. This list is a curated set of some of my public projects that I encourage you to checkout.
Not as easy to show, but I also help run a few of my university friend's websites and stacks via NixOS & Podman for ease-of-orchestration.
Current Focus
- Interpretable ML and specalised feature engineering
- Tooling for reproducible workflows (MLFlow/Optuna)
- Communicating technical work to non technical audiences
Contact
- GitHub: https://github.com/Scot-Survivor
- Email: questions@scot-survivor.com
Top Pieces
Transformer & Performer (from scratch)
Originally my A-level NEA project, I continue to work on this (dubbed "Gavin") throughout my academic experience as I learnt new tools and approaches, especially with recent advancements in Large Language Models.
An implementation of transformer and performer architectures from first principles to understand attention mechanics and performance tradeoffs.
- Focus: attention math, efficiency, training workflow
- Stack: Python, Keras (TensorFlow), C++ for data Processing + Speed
- Repo: https://github.com/Gavin-Development
Particle Swarm Optimisation Explainer
A Particle Swarm Optimisation visualiser that makes the algorithm's behavior easy to inspect, explain, and teach. Made during my years at university to assist in my own learning, and future students at University of Kent
- Focus: optimisation intuition, visualisation
- Stack: C++, ImGui & CMake
- Repo: https://github.com/Scot-Survivor/Better-PSO
Coming soon | Dissertation
Expect a write up of my dissertation, my thoughts and process of how I went about it. ETA: July/August 2026