About Me
I am a mathematically rigorous Data Scientist specialising in agentic AI systems and end-to-end machine learning pipelines. My work bridges theoretical mathematics, statistical learning, and practical business impact, consistently delivering functional autonomous systems within 8–12 weeks.
I have extensive experience with AWS and Google Cloud platforms, deploying scalable ML solutions using SageMaker, Vertex AI, and serverless frameworks. My focus is on optimising performance, reducing costs, and translating complex data and algorithms into tangible business results.
Key Areas of Expertise:
Agentic AI & autonomous system design
Statistical learning theory & optimisation
Monte Carlo simulations, DCF, and decision-theory frameworks
Risk algorithm design & scenario-based evaluation
Cloud-native ML architectures and pipeline deployment
MCMC model training & continuous retraining
Failure prediction, anomaly detection, and machine vision
Model validation, A/B testing, and performance monitoring
💻 Skills & Tech Stack
Python | Lean | Ruby | TensorFlow | AWS | Google Cloud | Microsoft Power BI | Git | Docker | LaTeX
📂 Projects
AMP SPC Project
A machine learning project focused on statistical process control and predictive analytics.
View ProjectMachine Learning MCMC Model
Full MCMC model training pipeline with continuous retraining and visualisations. Includes PDFs, images, and results of predictions vs. actual data.
View ProjectRisk Algorithm – Aircraft Parts
Analysis and risk prediction for aircraft parts using Python, JAX, and GARCH modelling, with formal mathematical proofs verified in Lean 4.
View ProjectBioinformatics – Squamous Cell Carcinoma
Analysis of squamous cell carcinoma using Python & Jupyter Notebook.
View ProjectLean Theorem Proving
Formal mathematics proofs using Lean 4 and Mathlib, covering number theory, algebra, and logical reasoning.
View ProjectSystolic BNN Accelerator (V2)
Redesigned BNN chip for Tiny Tapeout — XNOR dot product, systolic engine, signed bias, hardware feature expansion, and balanced popcount tree. Same 16 neurons, fraction of the silicon.
View Project16-Neuron Binary Neural Network
A full BNN inference layer in silicon on Tiny Tapeout — 16 binary perceptrons classifying an 8-bit input simultaneously in a single clock cycle. No CPU, no software, just logic gates.
View Projectτ²-bench – LLM Reliability Evaluation
LLM-vs-LLM agent evaluation across customer service domains, measuring Pass^k reliability scores.
View Project🏆 Work Experience Highlights
AJW Aerospace – Data Scientist
Apr 2025 – Present, Sussex
Designed and deployed autonomous agent AI systems for rapid business implementation
Developed risk algorithms and scenario-based evaluation models
Conducted Monte Carlo simulations, DCF analysis, and scenario-based risk assessments
Trained MCMC models with continuous retraining for adaptive decision-making
ASN Submarine Cables – Data Scientist & ML Engineer
Dec 2023 – Apr 2025, London
Converted corrective maintenance to predictive maintenance using ML
Developed failure prediction and anomaly detection pipelines using machine vision
Led Kaizen and Lean Manufacturing initiatives to improve KPIs and reduce defects
Applied DMAIC and 8D problem-solving for continuous process improvement
McLaren Racing – Complex Data Analysis
Mar 2020 – Dec 2021, Woking
Implemented SPC and predictive models to reduce manufacturing defects
Led FMEA and PPAP processes to ensure quality and reliability
Applied machine learning for failure prediction on manufacturing lines
📊 GitHub Stats & Badges
Commits (2026)
500+
across all repos
Lean Proofs
30+
machine-verified
Tiny Tapeout
2 chips
in silicon
Experience
10+ yrs
ML & data science
Profile Views
portfolio visits
🎨 Hobbies
Skiing, walking, cycling, theatre, golf