Co-founder & CEO, Kimova AI · Postdoctoral researcher, TalTech
AI researcher and founder.
I build Kimova AI, an agentic AI co-auditor for ISO, SOC 2 and privacy audits, and at TalTech I research how vision-language models perceive and reason. PhD from Brown University & ANITI.
What I do
Can we trust what an AI concludes? I work on it as a scientist and build on it as a founder.
Co-founder & CEO · Tallinn, since 2024
An agentic AI co-auditor for ISO, SOC 2 and privacy audits, for audit firms and in-house compliance teams.
Postdoctoral Researcher · TalTech
Do vision-language models reason, or only pattern-match? I build diagnostics and benchmarks that tell the two apart.
Selected research
Peer-reviewed at TMLR (J2C at NeurIPS 2026), CoNLL (oral at ACL 2026), ICLR and Neural Computation — research dimensions, work in progress and the full list on the research page.
Mohit Vaishnav, Tanel Tammet
Conference on Computational Natural Language Learning (CoNLL 2026), co-located with ACL 2026
Enrico Vompa, Tanel Tammet, Mohit Vaishnav
Transactions on Machine Learning Research (TMLR)
A linear probe on a VLM’s raw visual embeddings sets a diagnostic baseline that separates weak perception from weak reasoning on abstract tasks such as Bongard problems. Most models fall short of it; adding a contrastive objective to next-token training reorganises visual representations and lets models clearly surpass it.
Mohit Vaishnav, Tanel Tammet
arXiv preprint arXiv:2501.13620
Whether it's Kimova for your audits, a research collaboration, investment or speaking — I'd love to connect.