PhD researcher at Concordia University (Building Engineering) applying Machine Learning and AI to built-environment challenges. Founder of Le Laboratoire Kalkin, an AI consultancy specialising in multi-agent LLM systems, RAG pipelines, and financial forecasting. Speaker and organiser at The AI Collective, Montréal's applied AI community.
// AI Philosophy
Privacy-First AI
AI systems should augment human judgment without harvesting data. On-premises deployment and local LLMs aren't limitations — they're design choices.
Local-First Systems
Infrastructure that runs on your hardware keeps you in control. Cloud dependency is a liability; reproducibility and ownership matter.
RAG vs. Agents
RAG retrieves. Agents reason. Most real-world problems need both working in concert, with hard guardrails against hallucination at every node.
Why Most AI Products Fail
Over-promising, under-constraining, and ignoring domain knowledge. Good AI is 20% model and 80% pipeline, data quality, and human oversight.
Research on LLM-Agent Architecture – McGill University
Literature Review (May–Aug 2024): In-depth study of encoder/decoder transformer models, AI hallucination causes, detection, and mitigation via LLM agents and RAG systems.
Implementation (Jan 2025–Jun 2026): Prototyping multi-agent architectures in SmolAgents, LangChain, and AutoGen; evaluating performance and hallucination reduction.
Sep 2025 – Present
PhD Researcher – Concordia University · Building Engineering
Supervisor: Prof. Bruno Lee. Research on ML and AI applications for energy optimization in the built environment.
Collaborating with Volt Age on applied energy intelligence, bridging simulation-based building models with real-world deployment.
Sep 2025 – Ongoing
Speaker & Organizer – The AI Collective (AIC)
Speaker: Cracking the Formula for AI & ML — The AI Collective × McHacks & Plotly, McGill University; ML, LLMs, RAG & AI Agents — Simply Voting Inc.
Organiser: AIC × Deck Meetup (April 2026); The AI Collective MTL. Topics: LLMs, RAG, multi-agent systems, hallucination mitigation.
Mar 2021 – Feb 2022
Project Manager – Schlesinger Group
Delivered multiple B2B and B2C quantitative research projects, providing actionable insights.
Managed end-to-end research functions — data collection, processing, programming, and tabulation.