Mikoláš Belec on Data Talk: AI Adoption in Practice

Data Talk is a Czech and Slovak data community and podcast that features hands-on conversations with practitioners, founders, and teams building with data and AI.

In Data Talk #161, Siesta AI CEO Mikoláš Belec shares a grounded view on AI adoption: what breaks after the demo, what to standardize early, and how to scale usage without turning your company into a patchwork of disconnected tools.

Listen to the episode: https://www.buzzsprout.com/2034779/episodes/18218051

What the episode covers

The conversation connects three threads that matter if you are building or buying AI for a mid-market or enterprise environment.

Mikoláš shares his founder story, how he got into programming by chance, moved into hospitality, built a SaaS platform for hotel and Airbnb management, and successfully exited it. That path explains why his view on AI is operational: software only matters when it actually changes how work gets done.

They also get into the pace of AI development and the key question: how do you adopt new tech in a way that is structured enough to be safe, but not so slow that it kills momentum.

Finally, Mikoláš argues that the current wave of vertical AI tools does not make sense for many mid-sized companies, and that there is space for a platform approach, including a platform built from the Czech Republic.

The main takeaway

AI adoption should be treated like an operating model decision, not a tool install. You need clarity on where AI is allowed to access data, who owns outcomes, and what happens when the answer is wrong.

Non-obvious insight: the more “helpful” the tool looks in a demo, the more damage it can do at scale if governance is missing. A fast pilot can create a slow organization if every team ends up with a different workflow, different permissions, and different expectations.

Why it is still early

Mikoláš also reflects on time in San Francisco, where he reinforced a view that AI is still at the beginning. The market is expanding, which means it is not too late to build, invest, or adopt, but it does reward teams that learn quickly and ship in iterations.

If you want to explore what a platform-led rollout can look like, you can also book a demo: https://siesta.ai/demo