Articles
Starting from the end, and failing at building a startup
Cracking Tic-Tac-Toe!
[SPANISH] ¿Comenzó el colapso de la economía venezolana en 2014?
Reinforcement Learning basics: Stationaty and non-stationary multi-armed bandit problem
Evolving Deep Neural Networks
Projects
Lagrange: AI-driven Scheduling for Restaurants
Lagrange started as a staffing schedule generator for restaurants. The problem to solve is simple, schedule the right amount of staff at all times. This means forecasting how many customers will the restaurant receive per unit of time (the demand), how many customers each team configuration can handle (the capacity) and to use this information to build actual schedules ahead of time that supply enough team members in all areas to deliver to all customers while minimizing the payroll (the plan), all while respecting staff preferences and business rules (the constraints)
Auto-Mock: A more flexible Δ-MOCK implementation
MSc Data Science dissertation. Developed further the clustering algorithm Δ-MOCK, this version removes some sensible parameters while keeping performace. My advisors were Richard Allmendinger and Cameron Shand.
Playground
Curated collection of notebooks and code files I have worked on while learning a wide range of data science subfields, such as Reinforcement Learning, Natural Language Processing, Deep Neural Networks, Genetic Algorithms, etc. Some of these are accompanied by a pdf and/or article.
About me
Hello! I'm Luis Da Silva, originally from Venezuela and now based in the UK. My background is a blend of Economics and Data Science, but my career has taken an exciting turn towards Software Engineering and Entrepreneurship.
I'm the founder of Lagrange Systems, where we're using AI to revolutionize how restaurants manage pricing and staffing. Before that, I honed my technical skills as a Tech Lead at EMIS Health.
I thrive on challenges and I'm always on the lookout for intriguing problems to solve. And when I'm not immersed in code, you'll likely find me exploring the countryside on a run 😉.