Daniel Bacon

Developed a grounded AI and analytics capstone project using ball-by-ball cricket data across Test, ODI, and T20 formats. Built a Flask-based web application powered by a large language model to provide interactive, data-driven cricket insights. Conducted exploratory data analysis, created performance metrics, and implemented a chatbot interface to deliver accurate contextual responses to user queries. Focused on presenting complex statistical information clearly and effectively for non-technical users.

Built a full-stack automated trading system with live portfolio monitoring and strategy visualisation. Implemented a Flask web app to execute trades via the Alpaca API, with yFinance as a fallback. Integrated AWS S3 for caching data and logs, and scheduled background tasks with APScheduler. Developed interactive Plotly charts to track portfolio performance against the S&P 500 (SPY), implemented dynamic allocation algorithms based on analyst target prices, and deployed the application to Heroku.

Built a full-stack fitness tracking and coaching app with an adaptive weekly planning loop. Users log gym, cardio, physio, weight, and steps, and the AI generates structured weekly plans in JSON with actionable workout details (sets/reps/time). The planner evaluates adherence against prior plans, learns consistency patterns, and adjusts focus week-to-week while syncing sessions to FlowSavvy for calendar-based execution.

Completed a group project as part of the Digital Futures programme, analysing simulated company data to identify revenue growth drivers, customer acquisition trends, and geographic expansion opportunities. Applied Python and Tableau for data cleaning, exploration, and visualization. Developed actionable insights for non-technical stakeholders and delivered a presentation summarising findings and recommendations. Focused on combining quantitative analysis with clear business storytelling.

Conducted a solo data-driven project analysing over 5,000 films to explore the relationship between budget, revenue, genre, and profitability. Used Python for data cleaning, statistical analysis, and visualisation, creating metrics to identify trends and patterns driving box-office success. Built interactive charts to communicate findings and provided insight into factors separating blockbuster hits from financial misfires.

A data-driven project analysing customer churn for Swan Teleco. Conducted a full workflow: data cleaning, exploratory analysis, and predictive modelling using Random Forest, Logistic Regression, and SVM. Generated churn probability scores, identified the top 500 at-risk customers, and provided actionable insights to improve retention, such as targeted incentives and improved onboarding. Achieved 85.5% ROC AUC with the Random Forest model.

Analysed global health and socioeconomic factors influencing life expectancy (2000–2015) using WHO data. Built a complete data science pipeline: exploratory data analysis, feature engineering, and predictive modelling with linear regression. Deployed a live web interface allowing users to predict life expectancy using either an advanced or simplified model while maintaining ethical and privacy-conscious design.

Co‑founded and moderated a 2,500+ member Discord server for University of Southampton students — implemented moderation bots, hosted virtual events, and supported community engagement throughout the COVID‑19 period.