Computer Science graduate from Université de Montréal, working at the seam between production software engineering and applied ML. Software engineer at GoTo, research assistant at GEODES, and member of the Cohere Labs Open Science Community.

1st of 77 teams at the Bell x Mila x BuzzHPC x Kids Help Phone AI Safety Hackathon. Built a multilingual guardrail stacking a fine-tuned mmBERT classifier with a Cohere LLM judge using a five-question chain-of-thought framework. F1 0.906, recall 0.969. Presented at the CRIM 40th anniversary conference.

Research assistant at GEODES (software engineering lab at Université de Montréal, advised by Prof. Houari Sahraoui). Fine-tuning small open-source LLMs for automated code review with QLoRA / PEFT, and designing evaluation methodologies that benchmark generated reviews against human reviewers.

Software engineer at GoTo. Grown from three internships into an engineering role, shipping production code across frontend architecture (React/TypeScript), backend data pipelines (Java), and infrastructure / quality engineering.

Ubisoft GameLab Competition 2026 as gameplay programmer, representing UdeM. Unreal Engine 5.6, split-screen local multiplayer, player powers.

Languages: TypeScript, Python, Java, C++ Web: React, Node.js ML / NLP: PyTorch, Hugging Face Transformers, QLoRA / PEFT, BERT / mmBERT, LLM fine-tuning, LLM-as-judge evaluation, chain-of-thought prompting Game dev: Unreal Engine 5, Blueprints Infra: AWS, CI/CD, observability

LinkedIn · yanis.chalel@umontreal.ca

Open to Software Engineer and ML Engineer roles across Canada.