Staging environment
Free Lesson

Why Tool Calling Breaks—and What to Do Instead

30 min
Jun 2, 2025 12:00 PM
Virtual (Zoom)

In this video

What you'll learn

Why most “tool use” patterns break at scale

Understand the limitations of atomic function calls and ReAct-style orchestration in real-world systems.

How to design stateful, multi-turn business logic

Learn how to guide users through complex, multi-step processes—without relying on brittle prompt chaining.

How Process Calling works in production

See how to combine LLMs and deterministic processes to deliver robust, explainable conversational experiences.

Debug less and ship faster with modular flows

Discover how structured, reusable business processes reduce flakiness and accelerate iteration.

Why this topic matters

Most agent frameworks give LLMs tools and hope they pick the right one. But real conversations aren’t API-ready—users say things like “you messed up my order” and expect resolution. That requires branching logic, memory, and follow-up. In this lightning lesson, Alan Nichol and Hugo Bowne-Anderson show how Process Calling makes agents reliable, inspectable, and effective.

You'll learn from

Hugo Bowne-Anderson

Hugo Bowne-Anderson

Podcaster, Educator, DS & ML expert

Alan Nichol

Alan Nichol

CTO, Rasa, OSS Sorcerer

See all products from Hugo & Stefan