Developed an end-to-end MLOps pipeline using AWS services. The goal is to automate the deployment and monitoring of machine learning models.

Looking to develop a full-fledged application leveraging AWS services based on TermiCATor. The goal is to leverage serverless technologies, build a fully automated SageMaker pipeline, and more.

Developed a radiomics evaluation to predict outcome in locally advanced rectal cancer with tumor deposits. Analyzing RMI images and clinical data.

Developed an application to simulate IoT devices leading to alerts and notifications. Prioritizing scalability. Using Spark Streaming, Scala and Terraform for IaC. Kafka locally. Kinesiss on AWS.

Developed a weather prediction application using React, deployed using Azure. Users give image or tabular data, the model predicts the weather.

Developed a movie recommender system combining collaborative filtering and content-based filtering. Using MovieLens and IMDB datasets.

Categorizing 128x192 pixels images into 10 different classes. Using a CNN with less than 30 layers. Ranked 6th out of 99.

Implemented different models for classic and aspect-based sentiment analysis on restaurant reviews. N-grams, Naive Bayes, Logistic Regression, TF-IDF, word2vec, neural networks, FFNN, RNN, Transformers

Developed a financial data analysis tool inspired by TradingView. Storing data in a TimescaleDB database, using Dash for the frontend and Python for the backend. With indicators feature and more.

Developed during my internship at Université de Montréal. Flutter application with a FastAPI backend. Enabling users to easily create and optimize trampoline routines for gymnasts through bioptim (optimal control library).

Developed a first version of my portfolio website using React and Ant Design. Deployed with AWS Amplify.

Designed and developed a web-based IDE for the Shrek programming language. The IDE features video player, code editor, python interpreter, parent control, git versioning, maven and more.

Developed an application (with GUI) to simulate the optimization of snow removal in Montreal. Solution based on eulerian path, optimizing cost and time given a resource constraint.

Implemented a compiler for Andrew W. Appel's Tiger language using C++

Implemented a POSIX-compliant shell from scratch.

Developed an image processing script to segment coins in an image and recognize their value (no ML involved).

Developed a chess AI, back-tested by playing games on lichess.org using the API. Achived estimated 1800 ELO.

Developed an application to give users the ability to train and visualize machine learning models and bayesian networks on their data without writing code.

Developed a program for a LEGO MINDSTORMS robot guided by a color sensor and an ultrasonic sensor to retrieve pucks, leading to a 1m50s performance (9 pucks, 2mx3m play field).