At Lightbase Group, I’m developing an AI-driven energy audit platform that extracts data from Hydro-Québec and Énergir invoices, models energy consumption, and uses agentic AI to generate recommendations and structured audit reports.
I also designed and deployed a plant health monitoring pipeline, training computer vision models on labeled plant images and integrating multimodal LLMs to generate diagnostics and alerts.
For an IoT lighting control platform, I developed backend services and dashboards for monitoring and controlling connected devices. I also engineered AWS data pipelines and backend services to support AI model training and analytics.
At Concordia University, I developed anomaly detection algorithms for industrial datasets and researched unsupervised learning methods. Earlier, as a Machine Learning Developer Intern at the University of Manitoba, I optimized downlink precoding using graph neural networks and VAMP.
