MURTEC — the Multi-Unit Restaurant Technology Conference — is now in its 31st year, which tells you something about how long the restaurant industry has been taking technology seriously. More than 800 executives, operators, and solution providers gather at Caesars Palace in Las Vegas each March to compare notes, argue about what’s next, and close the kinds of deals that take months of email to arrange. Three of our partners attended as sponsors of the inaugural AI Summit at MURTEC, a full day of programming dedicated entirely to helping restaurant leaders understand AI not as a concept but as something to actually implement.
We went in with a hypothesis and came out with it mostly confirmed, with a few useful complications added. The hypothesis: most companies in this space know they need to move on AI, don’t quite know how to start, and are spending a lot of energy on the glamorous parts of the problem while the unglamorous parts wait. MURTEC didn’t disprove that. But it gave it texture.
Event: MURTEC 2026 · Location: Caesars Palace, Las Vegas · Dates: March 9–11, 2026 · Sirrus7 Role: AI Summit Sponsor
The keynote: the most honest thing we heard all week
The standout moment was the keynote from a former Chief AI Officer for enterprise, with 200-plus AI deployments behind her and a track record that includes helping launch one of the first major enterprise AI platforms in 2011. That depth of experience gave the keynote an earned authority that you don’t often get from conference stages. The central argument: companies are impatient about AI ROI, and it’s going to cost them.
The deployments that look glamorous — chatbots, generative content tools, consumer-facing AI — are often the hardest to measure. The ones that actually move the needle are in procurement, demand forecasting, and scheduling. She also drew a sharp distinction between using AI and doing AI — the former means you’ve plugged in a tool; the latter means you’ve built the infrastructure and judgment to measure whether it’s working. The warning she kept coming back to: don’t let convenience become dependency. Outsource your tasks, not your critical thinking.
“Don’t drive faster off a cliff. Outsource your tasks, not your critical thinking.”
The “Executive AI Compass” framework stages AI adoption across four modes — and argues most organizations are skipping straight to the hardest one before the foundational work is in place.
Micro-Tasker — AI handles specific, bounded, repeatable tasks. Low risk, fast ROI. The right place to start — and often undervalued.
Co-Pilot — AI assists human decision-making. The human remains in the loop and retains accountability. High value when done well.
Pilot — AI leads the process with human oversight. Requires mature data infrastructure and governance to be trustworthy.
Coworker — AI operates autonomously within defined boundaries. Most organizations aren’t ready for this — and that’s okay.
What we heard on the floor
As a sponsor of the AI Summit, we had a lot of conversations with people who were attending specifically because they knew they needed to figure this out and weren’t sure how. A few patterns came up repeatedly enough to feel like signal rather than noise.
The first is that excitement is abundant and direction is scarce. Companies know AI is important. They’re hearing it from their boards, from their competitors, from the conference agenda. What they don’t have is a clear first move. “Where do we start?” was probably the most common question we fielded, and it was asked genuinely rather than rhetorically. The answer — start with your data, before you worry about the AI layer on top of it — is not a glamorous answer, but it’s the right one.
The second pattern is that data readiness came up in almost every substantive conversation. Not “do we have data” — everyone has data — but “is our data in a state where AI can actually use it?” Clean, structured, accessible, consistently labeled data is the unglamorous precondition for everything else. Many organizations have years of transaction history, loyalty data, and operational data sitting in systems that don’t talk to each other, in formats that weren’t designed for machine learning. The AI strategy conversation is actually a data strategy conversation, and most companies need to hear that before they buy another platform.
The third pattern was tech stack modernization appearing alongside AI almost every time. You can’t bolt a sophisticated AI layer onto a twenty-year-old data architecture and expect it to perform. The organizations seeing real results from AI investments are the ones that have also done the work of modernizing the underlying systems — moving to cloud-native infrastructure, breaking down data silos, and establishing clean APIs between their operational platforms. That work is less exciting than a demo. It’s also what makes the demo possible in practice.
Our takeaways from MURTEC 2026
1. Start with non-glamour use cases. Procurement, scheduling, demand forecasting, and inventory management return faster, more measurable ROI than consumer-facing AI. Build your track record there before you aim for the spotlight.
2. Data readiness is the real conversation. Every company we talked to had data. Very few had data that was clean, accessible, and structured for AI use. Solve that problem first — the tools come after.
3. Measure effectiveness, not just efficiency. Going faster in the wrong direction is still going in the wrong direction. The right question isn’t “are we doing things faster?” It’s “are we doing the right things?”
4. Not all AI is the same. Vision, speech, predictive analytics, and generative AI solve different problems and require different infrastructure. Treating “AI” as a single category is how you end up buying the wrong tool for the problem you actually have.
5. Outsource tasks, not critical thinking. The most repeatable line from the keynote — and the one most worth writing on your whiteboard. AI should handle the tasks that consume your team’s time without requiring their judgment. The judgment is the part you keep.
6. Tech stack modernization and AI strategy are the same conversation. You cannot layer sophisticated AI on top of fragmented, legacy infrastructure and expect it to perform. Modernization isn’t a precursor to AI strategy — it is AI strategy.
What this means for where we’re going
None of what we heard at MURTEC surprised us, exactly. But it confirmed some things that are easy to second-guess when you’re in the middle of client work. The gap between organizational excitement about AI and organizational readiness for AI is real, it’s large, and it’s going to take time to close. The companies that close it will have done so by doing unglamorous things well — cleaning up their data, modernizing their stacks, building the internal muscle to evaluate AI output rather than just trust it.
That’s exactly the kind of work Sirrus7 was built to do. Not the demo. The part that comes before the demo, and the part that comes after it.
If you were at MURTEC and want to keep the conversation going, or if you weren’t there but the themes in this piece are ones you’re wrestling with inside your organization — we’d love to talk. The context window is always open.