Expedia, Global Hackathon 2024
Events Based AdCopy Content Generator
  • Prompt engineered and built ChatGPT 4.0 integration for an ad copy system for events-based travel in a team of 6 using Python and Go. Laid groundworks for a new Expedia marketing initiative.
Oct 2024
Expedia, Global Hackathon 2023
Mobile Ad Image Generator
  • Developed a personalised advert image generation engine in a team of 2 using ChatGPT 3.5 and DALL-E 2 models as an investigation to drive click through rate performances for Expedia’s websites.
Sept 2023
Expedia, Global Hackathon 2022
Multigroup Travel Bookings Scheduler
  • Achieved UK regional finalists in a team of 6 by using Expedia’s in-house front-end framework with React to build a room booking system for multi-grouped travellers with overlapping travel schedules.
Oct 2022
Swift, SIBOS Hackathon
Managing Risk and Financial Crime
  • Represented Capco in the “Managing risk & financial crime” track and developed a graph network visualiser with D3.js for analysts to detect potentially fraudulent transactions in cross-border payment journeys.
Oct 2020
Facebook, Online Hackathons
Wit AI Chatbot
  • Competed and built a messenger weather chatbot across 2 days using Flask, Heroku, SQL Alchemy and PostgreSQL.
  • Fast learnt how to use Wit-AI and Facebook messenger APIs.
Jun 2020
BlackRock, Algothon
Global Supply Chain Stock Price Predictor
  • Competed in a team financial themed hackathon to build a stock price predictor with global supply chain data and a deep multi-input layered neural network using Keras.
  • (Leveraging Quandl and Thomson Reuters APIs).
Oct 2019
Imperial College London
K-MNIST Challenge
  • Developed a 97.9% accurate deep convolutional neural network (CNN) to recognise ancient Japanese characters in a seven-day team project – a contender to state-of-the-art models.
  • Developed a PyTorch pipeline involving image augmentation and performance tools to fine-tune models.
May 2019
Imperial College London
Genetic Algorithms Challenge
  • Implemented a GA model to determine optimal mineral recovery systems using OpenMP and MPI parallelisation frameworks in C++ for a seven-day team project.
  • Analysed code implementation and run-times to improve parallelisation strategies.
Mar 2019