Autumn Projects

A collection of things I’ve built, explored, and learned from.

AI Governance Risk Tracker - Ongoing

Building a tool that scrapes and analyses corporate AI governance policies and guidelines to then score organisations on fairness, transparency and accountability. This will be done using Python and NLP.

Python · NLP · Web Scraping ·

UK Health Inequality & Life Expectancy Analysis

I analysed how life expectancy varies across UK local authorities by deprivation, education and demographic factors. I cleaned and merged ONS, IMD and Census 2021 datasets in SQL, engineered new indicators, and modelled correlations between deprivation scores and average life expectancy. Using Python and Power BI, I built an interactive dashboard highlighting geographic inequalities and patterns across England and Wales.

SQL · Python · Power BI · Census

NHS Pharmacy Dispensing Forecasting

I used publicly available NHS Open Prescribing data to analyse and forecast monthly dispensing volumes. I aggregated data in Python and engineered lag features, analysed seasonal trends, and built a forecasting pipeline comparing ARIMA and Prophet. Prophet performed best, capturing recurring monthly and winter peaks. I visualised projected volumes and emerging trends using Matplotlib and Plotly, demonstrating how prescribing data can support operational planning.

Python · ARIMA · Time Series

Recession Impact on UK Financial Markets

I examined how UK recession periods influence stock market behaviour, focusing on FTSE indices during historic downturns. Using Python, I cleaned ONS GDP data and Yahoo Finance pricing data, applied event-study methodology, and ran regression models to quantify the impact of recession shocks on returns and volatility. I visualised performance patterns across recession windows and identified the sectors most sensitive to macro-economic stress.

Python · Econometrics · Yahoo Finance ·

NHS Waiting Times Predictor

I modelled NHS treatment waiting times to understand where pressure builds across regions and specialties. After merging publicly available NHS datasets, I used regression and clustering to identify patterns in referral-to- treatment (RTT) times and detect high-pressure bottlenecks. I engineered features such as seasonal demand, capacity and specialty-level trends. The model grouped trusts with similar pressure profiles and flagged where waiting times were most likely to rise.

Python · Regression · Clustering

Sustainability & Emissions Performance Tracker

I created an Excel-based dashboard to analyse greenhouse gas emissions across UK local authorities using DEFRA and ONS datasets. I cleaned and normalised raw emissions data, calculated emissions per capita and per km², and built an interactive dashboard using advanced Excel functions, pivot tables and slicers. The tool highlights regional disparities in sustainability performance and identifies the highest- and lowest-emitting sectors.

Excel · ONS · Data Visualisation