Projects
I sometimes work on fun side projects. I’ve included a subset of these below.
systolic-array with AI helpers
4x4 output-stationary systolic array with MNIST inference pipeline in Verilog
remote_attestation with AI helpers
Simulates TPM-style attestation between a prover and verifier using Docker containers. - **Verifier**: Web app to define trusted file state and verify attestations - **Prover**: Measures local files, signs a quote, and submits it to the verifier
basic-shell
Super simple shell
minigrep
Quick implementation of grep in rust
stargate-timeline with AI helpers
Stargate Buildout Satellite Images Video
rfds with AI helpers
AI Security RFDs - Research ideas and open problems
Claude-Code-for-FPGA-Formal-Verification with AI helpers
Using Claude Code to develop formally verified network security hardware
match.py
A neural network framework built from scratch using artisanal (hand-written) cuda kernels.
hack_assembler
A basic assembler for the Hack ISA written in Python.
- physNet - Web app designed to help physiotherapists assess their patients Tensorflow.js, Computer Vision, p5, ml5, node.js
- What’s your problem? - Web app built for Upping your Elvis, Hackmed 2020 winner Tensorflow, flask
- th3g00d.space - Web app for gesture based passwords, HackSheff 2019 winner Tensorflow, Computer Vision, flask, Google Cloud
- Learning ASL using Transfer Learning - Web app for using transfer learning to recognise ASL for text based adventure game, HackCov 2019 winner Tensorflow.js, Computer Vision, flask
- Conquer the World \ - Native app for a multiplayer, geocaching inspired game where players compete to claim the most territory.
- Spinning Up in Deep RL
- \ - Worked through most of the curriculum (inclduing implementing PPO).
-
ML Approaches to Dementia Assessment - (peer-reviewed publication) supervised by Dennis Wang.
-
Agent-based modelling and multi-agent systems - highest mark in class.
-
On and Off Policy Approaches to Reinforcement Learning In Chess - highest mark in class.
-
Classification of the EMNIST Dataset via Competitive Learning
- Biologically Inspired Spiking Neural Networks for Signal Processing and Classification - undergraduate dissertation supervised by Professor Daniel Coca, 2nd highest mark in my department for my cohort.