
Stock Price Predictor
In this project, I attempt to combine my interests in ML and Finance in an attempt to predict the stock market.
I am a Computer Science and Math student at McGill University and currently serve as the Software Lead for McGill Robotics AUV. In this role, I specialize in state estimation, Docker containerization, and computer vision algorithm design. My current interests lie in perception systems, developing robust multi-sensor fusion techniques and leveraging geometric and statistical models to enhance scene understanding, reduce uncertainty, and effectively handle occlusions in autonomous navigation.
In this project, I attempt to combine my interests in ML and Finance in an attempt to predict the stock market.
Developed an LLM-guided exploration agent in ALFWorld using DSPy to infer a room’s occupant based on object attributes. Combined naive and advanced strategies with memory and reasoning to minimize steps and generalize across environments.
Developed Kiwi, an allergy-friendly Chrome extension that dynamically filters and highlights allergenic ingredients on Walmart’s website.
Conducted an in-depth analysis of distinguishing features between AI-generated and real images, examining contour properties, structural patterns, and frequency characteristics to reveal synthetic artifacts and enhance AI detection methods through statistical and visual analysis.
Built a predictive model for the Kaggle Titanic Competition, which involved determining passenger survival based on features like age, gender, and ticket class.
Developed a real-time lane detection pipeline in CPP for autonomous driving using classical computer vision techniques like Canny edge detection and Hough transforms.
If you want to get in touch to talk about work, projects, or maybe you'd like to work on projects together, feel free to reach out to me via email or connect with me on LinkedIn. I am always open to discussing new ideas, collaborations, and opportunities.