Skip to content
In PersonIn personPast event

From Big Model to Small Model: Distill Your Own Agent LLM

Tools & platforms

ClaudeGPT-5

Explore the topics

Event artwork for From Big Model to Small Model: Distill Your Own Agent LLM

This event has ended

This listing is kept as an archive. Explore upcoming events to find your next one.

See upcoming events

Supporting this event

Sponsors

Tech 42
About this event

What to expect

Running AI agents at scale is expensive. Every API call to Claude, GPT, or other cloud-based large language models racks up per-token costs, sends your data to third-party servers, and locks you into external infrastructure. This hands-on session shows you how to break free by distilling your own lightweight AI agent that runs entirely on your own hardware.

Hosted by the Montreal AWS User Group, this practical workshop walks you through the complete pipeline for building production-ready AI agents without recurring API fees. You'll learn how to capture high-quality conversational data from large models like Claude or ChatGPT, use that data to fine-tune a smaller open-weight model through knowledge distillation, quantize it for efficient inference on modest hardware, and deploy a fully functional agent that never phones home.

What You'll Learn

  • Knowledge Distillation Fundamentals: How to transfer capabilities from large cloud models to compact open-weight alternatives
  • Data Collection & Curation: Techniques for capturing and structuring training data from production conversations
  • Model Fine-Tuning Pipeline: Practical steps for adapting smaller models to your specific use case
  • Quantization & Optimization: Making your model run efficiently on local infrastructure
  • Agent Deployment: Building production agents on top of your custom model with no external dependencies

This session is ideal for ML engineers, DevOps professionals, and technical leaders who are tired of per-token pricing and want full control over their AI infrastructure. Whether you're building customer service bots, internal automation tools, or domain-specific assistants, you'll leave with a clear roadmap for deploying cost-effective, privacy-preserving AI agents.

The evening begins at 6:00 PM with networking, followed by a live presentation and demo at 6:35 PM. The session concludes at 8:00 PM after Q&A and feedback. The event takes place at the Trottier Building on McGill University's downtown campus.

Sponsored by Tech 42, an AWS partner specializing in AI/ML adoption, this meetup brings together Montreal's cloud and AI community for practical, production-focused learning.

[Register for this event](https://www.meetup.com/montreal-aws-users-united/events/313618109/?recId=ab04ff61-cf6f-4533-84cf-aa096e99637e&recSource=event-search&searchId=836f00ad-8d60-4817-8601-d99aa6ab7760&eventOrigin=find_page

Running an AI event?

List it here, free.

Meetups, workshops and conferences go in front of thousands of builders after a quick review.

List your event