In this workshop
AI agents are showing up in every product, but most of them are limited to responding with their training data, information they find from the web, and information they get from tool calls you’ve built to interact with your app. They don’t have native access to the other apps and services your customers use.
In this workshop, Taylor Reece shows how the MCP flow server turns the integrations you already have into tools AI agents can call: deterministic flows with production-grade auth and multi-tenant isolation, so agents built by you or your customers can take real action across any connected system.
Walk away knowing how to
- Build an agentic flow with an invocation schema so agents know what it does and how to call it
- Enable flows as MCP tools and connect any MCP-capable agent (or your in-product chatbot) to your integration's MCP endpoint
- Build agent flows in low-code or code-first, on the flows you already run in production
- Extend your product safely: deterministic execution, real auth, and per-customer isolation at scale
