// BLOG · AI HIRING

AI ATS: why we built heyree around real recruiting

By David Říha · Updated

Before cofounding heyree, I led hiring at FTMO, Vilgain, CloudTalk and Groupon. The problem was not simply processing more CVs. It was keeping track of the right people when spreadsheets, lost emails and disconnected tools got in the way. That experience is the starting point for how I think about an AI ATS: less routine work, more attention for candidates, and decisions that stay with people.

The frustration that became heyree

Our About us page describes the beginning plainly: recruiting tools made us think like machines, while candidates slipped through the cracks. We founded heyree in September 2025 and built it from Prague, initially for the Czech and Slovak market. My role is cofounder and CEO; Jiří Jaroš is our CTO and Adrián Horváth our AI Architect. That combination matters: recruiting experience defines the problem, while engineering and AI turn it into a usable product.

What an AI ATS actually is

An applicant tracking system keeps applications, candidate records and hiring stages together. An AI ATS adds assistance inside that workflow. Depending on the product, that can mean preparing summaries, suggesting matches or helping with communication. Those examples are not a universal feature list. heyree's About us page specifically describes Ree, CV scoring and candidate verification, alongside a shared hiring data model. Ask every vendor to demonstrate what its own AI does rather than assuming that the label tells you.

The goal is not to overlook the right person

Our public principles are speed without losing the human touch, respect for every applicant, clarity instead of tool chaos, and AI advising while people decide. For me, that is a more useful product direction than promising to process the largest number of CVs. In a trial, I would check whether the team can find the original application, see what happens next and give the candidate a clear answer. A polished score is not a substitute for those basics.

EU rules depend on the AI's intended use

Annex III, point 4(a), of the EU AI Act covers systems intended for recruitment or selection, including targeted job advertising, filtering applications and evaluating candidates. These uses generally fall within the high risk category. Article 6(3) provides a limited exception where the system poses no significant risk, including by not materially influencing decisions, and meets a listed condition such as a narrow procedural or preparatory task. Profiling natural persons remains high risk in this context. Having a recruiter make the final decision does not, by itself, remove the classification.

A real person must be able to question the output

The high risk framework requires effective human oversight, not a ceremonial approval click. Article 14 addresses understanding limitations, recognising automation bias and being able to disregard, override or stop the system. Under Article 26, employers using a high risk system have their own responsibilities, including assigning competent oversight and following instructions. Ask to see the original evidence behind a recommendation and what happens when the recruiter disagrees. These are evaluation questions, not a claim that heyree has received a compliance certification.

Data protection and the current timeline

heyree's About us page states that data is kept solely in the EU and is never sold to third parties. That is a company statement, not proof of GDPR compliance on its own: lawful processing, retention, access, candidate information and security still need checking. GDPR Article 22 restricts decisions based solely on automated processing that have legal or similarly significant effects, subject to defined exceptions and safeguards. Regulation (EU) 2026/1744 moved the application of the Annex III high risk rules to 2 December 2027. GDPR already applies, and existing AI Act prohibitions should not be confused with that later date.

How I would evaluate an AI ATS

Start with one role and a clearly defined workflow. Compare missed applications, time spent on administration, the evidence supporting the shortlist and how reliably candidates receive answers. Use candidate data only with an appropriate legal basis and safeguards. Ask the vendor for its intended use, risk classification and documentation. Choose the AI ATS that helps your team stay organised and accountable, not the one with the most impressive automation claim.

Sources and further reading

The heyree story is based on its public About us page. EU rules are summarised as checked on 13 October 2026. This is general information, not legal advice or a compliance certification of heyree.

// BUILDING AN AI RECRUITMENT PRODUCT?

I build websites and brands for AI ATS and AI hiring products. Email me at growthjungle@gmail.com.