HITEC 2026, day three: eight startups, and our first closed-door round table

Day three was our second day on the floor, and our day one and two recap is here. It started with eight founders on the E20X stage, each with four minutes to pitch, and ended with us in a closed room where hotel CIOs and suppliers talked openly about where all of this is going. In between were the booth conversations. More than once, the honest answer to a question was just "we don't know yet."

E20X: eight startups, four minutes each

Before the floor opened we sat in on E20X, HFTP's startup pitch competition, which this year put eight companies on stage in front of a judging panel and a packed room. The format is strict: four minutes to pitch, four minutes of questions, and the mic gets cut the moment the clock runs out. Lyle Worthington, who also moderated the Dark Side panel earlier in the week, ran the room and kept everyone calm and on time, which matters when you've got four minutes to explain your company, several of them pitching in their second or third language. Two prizes were on the line, the Judges' Choice worth 5,000 dollars and a People's Choice voted by attendees in the app. Here's each team, and what they pitched.

ABRA is a guest-data and personalization platform that pulls a full guest picture from every system and adds the things staff overhear, so a bellman can log that a child in room 212 misses her puppy and a stuffed animal arrives at the door. Co-founder Josh built it with his wife Dana, who ran operations for Four Seasons, Peninsula and Relais & Châteaux, with the goal of giving hotels one place where all that guest data finally lives.

Altek AI is building software to handle routine hotel operations on its own, starting with guest email: it reads an incoming request, works across the PMS and booking systems, and finishes the task before drafting a reply, turning ten minutes of jumping between systems into about thirty seconds. The Oslo team, led by co-founder Jon-Fredrik Hopland, says it's live in dozens of Nordic hotels and growing fast.

Bonafide works on getting hotels found correctly in AI search by pulling their scattered data on rooms, rates and policies into one clean, trusted source, so assistants like ChatGPT and Gemini answer accurately instead of guessing. Founder Tom Underwood says hotels typically start with about a third of their facts misaligned, and that cleaning it up lifts them past 80 percent correct.

Caleta builds what founder Daniel Hersey calls a nervous system for a property, pulling from the PMS, labor and work-order systems to catch problems before the guest feels them, for example spotting that a VIP's room shares an HVAC zone with two rooms reporting faults, then moving her and sending an engineer ahead of arrival.

Conscious Creative Group makes custom concierge characters, voice and text, with real voice actors and 30 languages, that know the property and the neighborhood, upsell and book amenities, and turn those guest conversations into useful data for the hotel. Founder Andrew Hummel, a former television showrunner, built "Ames" for the Palmer San Antonio as the HITEC demo.

Lobby automates the complex 70 percent of reservations most tools skip, the group requests, contracts, modifications and invoicing that arrive by email and PDF, sitting between the inbox and the PMS so an agent approves and Lobby does the rest. It's named for Daniela, the long-serving reservation manager at the founder's family hotel in Liechtenstein.

OptyLink is a patent-pending platform that automates first-time device installation, so set-top boxes, thermostats and locks configure themselves on power-on, and each configured device becomes a "scout" that helps onboard the next. Co-founder and CEO Nathan Nibras pitched it as the end of room-by-room manual setup, starting with Dish set-top boxes.

StayFull is an AI-native property management system aimed at independents, replacing the whole stack with one virtual agent fronting a team of specialist agents for revenue, marketing and operations, starting at 199 dollars a month against the 1,400 to 2,600 the founder says she used to pay. She runs her own four boutique hotels on it today.

At the booths

The booth conversations kept coming back to the same question from the first two days: who wins the booking, and who pays for it.

Oracle, the largest PMS company at the show, already has some agents running in hotels. Tanya Pratt showed a room-assignment agent that reads free text, so a travel agent's note that a guest is on honeymoon and should be upgraded gets picked up and acted on instead of sitting unread in a comment field. There's also a support agent that answers a question like how do I undo a checkout in plain language, and a check-in that runs itself. Oracle has built around 50 of these, with a couple already live. What she was less sure about was the payoff. Cloud had a clear value you could sell, she said, and AI doesn't yet, because no one has the results that would show everyone else how to do it.

Fareed Ahmad of HotelKey wanted to separate what's actually AI from what's just rules. Payment automation, he said, is rule-based and shouldn't be sold as AI, while deciding whether a returning guest with a past service issue deserves an upgrade is the real thing. He splits guest recognition the same way: the CRM and loyalty side decides what a guest should get, and the PMS carries it out, because, as he put it, AI isn't going to move the water bottles. HotelKey is running this year's AI as free pilots, since in his view you shouldn't charge for value you haven't proven.

Frank Trampert, nine weeks into Revinate, talked about why reviews matter so much for AI search. The models distrust a brand's own marketing and lean on guest reviews, because that's the real experience a customer shares in public. That makes the guest data Revinate has gathered for years useful for getting found by AI. He also said AI has to deliver a measurable result, and that the industry is over-promising while the proof is still thin. And he pointed to a public site the company launched to show its data practices, which not many do.

Natalie Kimball of Shiji was the most skeptical person we talked to, and she'd rather talk about the industry than her own product. Her view is that the OTAs have already won the AI distribution fight the same way they won search, because hotels can't pay enough to stay visible and missed the short window they had to act. She expects OTAs to add an agentic commission tier, maybe 20 percent against 15 for mobile and desktop, the way packaged bookings already cost more. The one hopeful thing she said was that guest reviews become the content that matters most with AI, because across thousands of reviews you can see consistent patterns and turn them into something bookable. It was the opposite of what most people on the floor were saying, and it stuck with me going into the round table that evening.

A first for us: the HN Round Table

After the floor shut, we hosted our first HN Round Table, something we'd been planning for the past month. For 90 minutes we brought six CTOs and CIOs from major hotel groups together with six senior supplier executives, with Floor Bleeker leading the discussion. No presentations, no sales pitches, just a conversation about where the enterprise PMS stack is heading. Those two groups rarely sit down together like this, and getting all of them in one room took a lot of effort.

It worked. People said what they actually thought. They agreed on some things, disagreed on others, and everyone seemed glad to talk openly without anyone trying to sell something. Because it was a closed-door session, we won't say who was there or who said what. But there was plenty in it, so we'll write a separate piece on what came out. And we're already looking at doing it again.

Where this leaves me

If day one and two left me excited and uneasy, day three reminded me that the best part of a show like this isn't the demos. It's the moments when people stop selling and tell you what they actually think. We got a lot of that today, on the E20X stage, at the booths, and especially in that closed room at the end.