Katja Gorlinski
Bio
Updated 07/28/26 · Provided by member · VerifiedFounder of iEVA (Finland), a platform that trains decision-making under pressure, grown from the iEVA Method and seven years of hands-on work with founders, managers and teams. I came to AI safety through that work. In 2019 I formulated a working hypothesis: the quality of data and environment loaded into a decision-making agent, whether a person learning cognitive skills or a model in training, essentially determines the safety of the systems that agent operates. The field has since turned to sociotechnical alignment; my hypothesis found itself in good company. This fall I pilot the human half with a real company team: baseline, behavioral metrics, pre-registered endpoints, change tracked over time. But this project tests the machine half of the same hypothesis: does curated data change a model's dispositions? I build solo, in the EU, and I believe in claims that can be falsified. Why fund me? I am a professional risk manager, and risk is the native language of AI safety. The method I mentioned is not a concept. I turned it into a working diagnostics product, from method to running code, and it has already survived contact with paying reality, a harsher filter than peer review. I am not from the ML industry, and the experiment is designed so that this does not matter: every choice is frozen before the first run, the analysis is fixed in advance, and a null result gets published under my name. As far as I can see, independence in this field is my strength, not a weakness, because a researcher with no institution behind her cannot afford a single unverifiable claim, and I build everything accordingly.
Links
Updated 07/12/26 · Provided by member · Verified- Personal Website
- https://www.i-eva.fi/
Projects
Grants
Updated 07/28/26 · By grantmaking.aiNo grants recorded.