See you in Chicago — plus giving your AI agents the story behind every screenshot

Our newsletter from August 20th, 2026. Feel free to sign up to get this directly in your inbox

We're heading to Chicago for the Support Driven Expo!

It's almost here! We'll be in Chicago from August 24–26 for the Support Driven Expo — and we can't wait. Three days of conversations, practitioner-led sessions, and plenty of opportunities to connect with the people thinking about where customer support goes next.

AI shows up everywhere on this year's agenda, and for good reason — support teams are navigating rapid advances in AI, rising customer expectations, and growing pressure to operate more strategically. And it's a conversation we're deep in ourselves, which is why we're especially excited to give folks in Chicago an early look at what we've been working on.

Connect your AI agents to a living visual knowledge graph of your product

AI agents are getting much better at reading documentation, reasoning across sources, and even taking action on a user's behalf. They can read the screenshot already sitting in your help articles — but what they can't see is whether that screenshot still matches your product, what's changed since it was captured, or how it connects across your entire documentation.

That's exactly what we're starting to expose with LaunchBrightly's upcoming MCP Server.

We're building an MCP server that connects your AI agents to LaunchBrightly, giving them access to every screenshot, every article, and the relationships between them — a living visual knowledge graph of what's actually true about your product. The first version is read-oriented, letting agents query that knowledge graph using the same underlying data you see inside LaunchBrightly today. For example, you'll be able to ask your AI agent things like:

• Show me the latest screenshots associated with the Settings section.
• Which screenshots changed in Sunday's Help Center Audit?
• Which screenshots are currently out of sync?
• Show me only the changed screenshots from the Billing section.

Instead of your AI seeing screenshots as static pixels, it can understand the product workflows behind them — where they appear, what they represent, and how they connect across your documentation.

And this is just the start! As we expand the MCP server, agents will move from understanding your LaunchBrightly data to acting on it directly — running audits, regenerating screenshots, assigning elements for review, accepting changes, and eventually building new screenshot automations themselves. To infinity and beyond!

Get a sneak peek

We're targeting a release for this first version later this Fall, but we've already got an MVP up and running and we'd love to show it off.

If you're in Chicago, come find us at the booth and let's geek out about it together. We'll show you what we're building, walk through the kinds of questions an agent can already answer, and talk about where we're taking it next. Or, if you'd like to talk screenshot automation more broadly, knowledge management, or simply say hi — we'd love to see you 🙂

Can't make it to Chicago? No problem. Grab some time on our calendar and we'll give you a sneak peek at the MCP server and what's coming next.