The Glassbox Effect - Why Enterprises Can't Ship What They Can't See
Most Claude agent demos work beautifully on stage and stall in production. The gap isn't the model — it's governance, observability, and knowing when an agent is the wrong tool entirely.
Jennifer Kiunke (Make) walks through the automation spectrum, from deterministic workflows to fully autonomous Claude agents, and shows where each belongs. You'll see how Make gives your Claude agents access to enterprise tools and knowledge through MCP while keeping every step visible, auditable and reversible — including how Claude Managed Agents surface inside Make Grid, so you can watch what your agents are actually doing at runtime instead of inferring it from the output.
What you'll take away:
When to use a deterministic workflow, when to use a Claude agent, and how to tell the difference before you build
What "glassbox" execution looks like in practice — and why black-box agents die in procurement
How Claude Managed Agents are monitored, traced and debugged in Make Grid
Real production use cases across sales, marketing and internal operations
How guardrails and human-in-the-loop approval get designed in, not bolted on
Who this is for:
Claude users evaluating agents for their organisation — founders, AI leads, heads of operations and anyone who has to answer for what an agent did. No Make experience required.
Every attendee receives 25,000 Make credits.
Full Claude Community House program on claudebcn.com