Madrid gives you two very different ways to spend an AI trip in March 2027. One puts you among hospitality operators deciding what technology to buy; the other puts you in a university auditorium working through how generative models actually behave.
I would not pretend those rooms are interchangeable. My Madrid shortlist contains two events, and the right one depends less on how often its programme says ‘AI’ than on the problem you need to solve when you get home.
1. HIP Horeca Professional Expo 2027: where AI meets the hospitality buyer
1–3 MARCH 2027 | IFEMA MADRID, SPAIN
HIP is a hospitality exhibition with a large congress attached, rather than a general-purpose AI conference. Its specialist summits include digitalisation, AI and robotisation, while its exhibition covers the technology and equipment used by hotels, restaurants and catering businesses.
That distinction matters if you sell AI. A restaurant group may care less about your model architecture than whether your system shortens ordering time, reduces waste or works with its existing point-of-sale setup. I would bring a working demonstration and a plain-language account of integration, staff training and costs.
If you run hospitality operations, reverse the conversation. Ask suppliers to show their performance on a busy shift, explain what data they need, and identify who takes over when automation fails. A polished demo is useful; a credible fallback plan is a buying signal.
HIP also deserves a place beside my guide to hotel technology events. If your AI product depends on reservations, pricing, property systems or guest service, a sector event can put you closer to the people who will live with the implementation.
2. GeMSS 2027: five days inside generative modeling
15–19 MARCH 2027 | UNIVERSIDAD CARLOS III DE MADRID, PUERTA DE TOLEDO CAMPUS
GeMSS, the Generative Modeling Spring School, is built for people who want to understand the methods beneath generative AI. Its stated scope runs from latent-variable and autoregressive models to diffusion and flow-based approaches; the published outline begins with lectures and practical sessions before moving to invited research talks.
I would put it high on a PhD student's list if generative modeling is central to a dissertation, or may become central soon. The smaller, teaching-led setting should make it easier to expose a weak assumption in your research than a broad expo would. Bring one stubborn technical question and enough background to discuss it.
Postdocs and research engineers can also use the week to check whether a method is ready for their domain, from scientific data to language systems. But if your immediate goal is procurement or meeting hospitality customers, GeMSS is the wrong room. Its value is depth, not a sales pipeline.
The organizers describe an application package that includes a research summary and CV, with a supervisor confirmation for students. Check the official page for the deadline and final programme before committing to travel; the timetable is marked tentative.
Two Madrid trips, two different returns
HIP and GeMSS are only two weeks apart, but I would judge them with different measures. At HIP, success might mean three serious buyer conversations, a shortlist of integrations or a pilot with a hotel group. At GeMSS, it might mean a better experiment, a useful critique or a collaborator who knows why your model fails.
That difference should shape your preparation. For HIP, write down the workflow, budget owner and operational metric you want to improve; then test exhibitor claims against those constraints. For GeMSS, read the recommended material, prepare a concise explanation of your research and identify the method you most need to understand.
There is a sensible case for both if you work on generative AI for hospitality, but ask whether you can turn the learning from one into a concrete question for the other. Two trips without a bridge between them are an expensive way to collect lanyards.
Which Madrid AI event fits your work?
PhD students and applied research scientists should start with GeMSS, especially if they need to compare generative-model families or sharpen a paper. Senior ML engineers should consider it when the hard part is methodology; they should choose HIP only when they need to learn the operating constraints of hospitality customers.
Hotel leaders, restaurant operators, hospitality product managers and AI founders selling into the sector should start with HIP. Go with a narrow use case and ask for evidence from a real deployment. If your role spans research and commercialisation, decide which gap is currently blocking progress and spend your travel budget there first.
If neither event matches your brief, browse my wider Spain AI events guide for research and industrial options in Barcelona. A better event in another city is worth the extra train ride if it introduces you to the right technical peers or buyers.
Did you not find the event you were looking for?
Tell yourself what you actually need: a specialist method, a buyer, a sector partner or a place to present work. Search for that outcome, then verify dates, programme and attendance requirements with the organiser before you book.