AI Mission Control

September 17, 2026

6:00 PM – 8:30 PM
Social House Roswell, 1098 Green St, Roswell, GA 30075

You attend AI events, act on what you learn, and your AI program still struggles. Eighteen months ago, someone greenlit a shiny new chatbot or LLM integration. Today, it’s either still not live, live and completely ignored by your users, or live and quietly making your internal processes worse. Sound familiar? That’s the reality for most enterprise AI programs right now. Maybe even yours.

This isn’t just another AI meeting. While so many people are busy hyping the next big breakthrough, not enough are actually talking about why these costly initiatives aren’t working. Instead of dodging the uncomfortable pitfalls, like garbage data, token cost overruns, and user adoption flatlines, we are diving straight into them. This is a highly collaborative, interactive event where we pull back the curtain, identify where things are getting stuck, share our professional missteps in a safe space, and work collectively to determine how to get these projects moving again.

Five stations, five station pilots, each exploring a different dimension of AI implementation:

Sky City: For infrastructure and third-party tools that were never integrated or adapted for AI. Pilot: Dorren Schmitt, PhD, VP of IT Strategy and Innovation, The Weather Channel and Allen Media Group.

Swamp Planet: For the promising program sinking in technical debt and bad data. Sponsored by LogicSpree.

Ice Planet: For AI initiatives that struggle to deliver measurable business value. Pilot: Mat Mathews, Chief Information Officer, Fortrex. Sponsored by Griffin.

Snow Monster Cave: For the evolving software development practices required in the AI era. Pilot: Mike Park, Chief Information Officer, Infor. Sponsored by Allata.

Asteroid Field: For navigating the growing demands of compliance, legal, and security. Pilot: Dante Jackson, Founder and CEO, Serket-Tech Security.

Real people. Real failures. Real solutions.

Capped at 125 attendees, each station is a focused group of about 25. You’ll be close enough to actually ask your specific questions, share your own war stories, and actively problem-solve. Every station is a guided discussion, not a talk: the pilot puts a short list of pointed questions to the room, and the room answers. Choose your top station in advance, and you’ll receive your final assignment at check-in to ensure all groups remain balanced and intimate.

Then comes the part nobody in any single station gets to see. When the stations wrap, everyone reassembles for the Throne Room, a closeout session built in real time from what all five rooms actually said. Each station is recorded and transcribed, and in the few minutes it takes to walk back to your seat, the five transcripts are synthesized into a single set of talking points: what each room found, and the patterns that only show up when you can see all five at once. John Trainor, President of Four Technologies, hosts and presents it live.

Everything about how the night works is public, before and after. The full method is on GitHub at https://github.com/atpconnect/atp-ai-bad-feeling: the question list for every station, the runbooks, the tool that builds the closeout from the transcripts, and the rehearsal harness we used to test the whole evening, including every way it could fail. After the event, the station transcripts and the closeout deck are posted there too, so you leave with the other four rooms’ findings and not just your own. Scan the QR code at the event or bookmark the link now.

This isn’t 125 people quietly watching five talking heads repeat the same safe, sanitized answers from a distant stage. This isn’t a panel. It’s the honest, hands-on debrief more companies should be running.

Event Leads