Optimization goes agentic: What three weeks of Custom APIs taught us

On October 21, we take the first step toward letting you build your own scheduling solution. You describe your scheduling problem in plain language. We turn it into a Custom API that plans around your rules and priorities.

Three weeks after launch, Tom and Lars look back at how it went. They’ll cover which problems people brought, where the process worked well and where it got stuck. They’ll also show what the first Custom APIs look like in practice.

What you’ll see

• The first Custom APIs. Real scheduling problems submitted since launch, and the APIs we built from them.
• What surprised us. Patterns in the problems people described, and what they taught us about turning a description into a working optimization solution.
• What we’re changing next. The improvements we’re making after two weeks of real use.
• Your questions, answered. Ask Tom and Lars anything about AI-powered scheduling, optimization or building your own Custom API.

Who it’s for

If you joined the launch webinar, this is the follow-up with real results. If you’re new to Timefold, you’ll see the whole path from a plain-language description to a working API, with real examples.