Build a Product Discovery Agent

Session Resources

Helping customers find the right product is one of retail's oldest and highest-leverage challenges — and one where AI agents can have immediate impact. Traditional search and filtering tools rely on exact keyword matches and predefined categories, leaving shoppers frustrated when they can't articulate precisely what they want. AI agents that understand intent, ask clarifying questions, and reason across a product catalogue can fundamentally change the discovery experience, increasing conversion and reducing abandonment.

In this code-along webinar, Ivan Bykanov, Product Lead at Lightspeed Commerce, will show you how to build a product discovery agent from scratch. You'll learn how to structure and prepare product data so agents can reason over it effectively, design an agent that interprets shopper intent, and deploy a working tool that guides customers to the right product.

Key Takeaways

• Learn how to structure and prepare retail product data for use in agentic AI workflows.
• Build a product discovery agent that understands customer intent and navigates large product catalogues.
• Explore real-world design patterns for deploying AI agents in retail and e-commerce settings.