Productionizing Agents : An Agent Factory Approach

About

In this talk, we’ll walk through how we approach agent development at Tiger - from the first idea to getting agents into production. We’ll cover what it takes to build and manage agents at scale, and share some of the practical things we’ve learned along the way.

Expect a deep dive into our Agent Platform - how we think about agent architectures, context engineering, observability, and more. Most importantly, we’ll talk about how a data flywheel mindset has helped us move faster, improve agent behavior, and make better decisions at every stage.

Key Takeaways:

  • Start-to-finish agent lifecycle: What it looks like to go from idea to production in a platform setting.

  • Why context is king: How we manage dynamic, high-quality context that actually helps agents perform.

  • The data flywheel effect: How feedback loops across usage, logs, and human signals improve everything — fast.

  • Observability isn't optional: Why metrics, traces, and debugging tools matter more than ever for agents.

  • What we wish we knew earlier: Honest lessons, mistakes, and trade-offs from building in the real world.

Speaker

Book Tickets
Download Brochure

Download agenda