William Fuller

Swile — Fintech company providing an employee benefits platform and card, enabling companies to distribute benefits across France and Brazil.

• Built internal ordering, invoicing and accounting systems, integrated SEPA payments on Swile-issued bank accounts, and developed data pipelines covering ingestion, data contracts and projections
• Drove the transition from third-party BaaS to Swile-issued bank accounts for invoice and order payments, taking full ownership from SEPA integration through to automated invoicing and accounting workflows

Tools : Ruby/Rails, PostgreSQL, Kubernetes, DBT, Redis, Kafka

Super G — ERP solution for beer brewers and winemakers offering tools for production, inventory, invoicing and payment management.

• Founding Engineer recruited to drive and accelerate end-to-end development, enhancing the product offering. Achieving 80% growth in number of active users and doubling revenue during first year.
• Led multiple projects from initial conception to implementation and delivery, aligning solutions with company goals and client needs. Notably developed invoicing platform and payment handling systems.

Tools : Go, PostgreSQL, React, Nginx

Instem / Notocord — Software suite of biological analyser used for pre-clinical trials in drug development.

• Responsible for prototyping and researching real time machine learning and deep learning models for biological analysis as well as the integration of these models in production.
• Conducted advanced research to create biological analyzers targeted at sleep scoring, using deep learning and machine learning methods to support drug research and development. Notably developed a sleep scoring algorithm achieving over 95% accuracy rate in rodents and smaller animals.

Tools : C#, Python, Matlab, Pytorch, ONNX

Tictactrip — Bus and train ticket purchasing platform offering all in one tickets combining buses and trains from different providers.

• Built and automated data-cleaning tools streamlining data processing, significantly improving search accuracy and system performance, while reducing computational costs. Halving the time needed to integrate new providers and generate internal stations graph.
• Took full ownership of all data processing tasks relevant to integration of new provider data. Enhancing the platform's efficiency and expanding its service offerings.
• Oversaw an intern, offering feedback and training to enhance their skills and knowledge.

Tools : Python, NodeJs, PostgreSQL, Docker

Imperial College London

Adaptive Signal Processing and Machine Intelligence, Digital Image Processing, Pattern Recognition, Probability and Stochastic Processes

Grade : Merit

University of Kent

Engineering Mathematics, Signals and Systems, Embedded Computer Systems, Digital Signal Processing and Control

Grade : First class distinction Award : IET Prize 2018

Rss reader — code

Rss aggregator written entirely from scratch using only the Go standard library. Deployed as a web app, running locally on a raspberry pi.

Tools : Go, Nginx

Nearby Velib finder — code

One click app to find the nearest Velib stations while on the move. Super simple app written in vanilla HTML/CSS/Javascript, accompanied by a caching server in Go regularly polling the Velib API.

Tools : Go, Nginx, Javascript

Solving Wordle with information theory — code

The goal here was to build a bot capable of solving the nytimes Wordle puzzle using only information theory. This bot finds the solution on average in 3.12 tries.

Tools : Python, Selenium