About 8Hats Lab
We are an independent research lab making AI-nativity a measurable, reliability-grounded discipline. Organizations of humans and agents are becoming the default; almost nobody can measure whether theirs is actually becoming one. We build the map, the scale, and the instruments — and we apply them to ourselves first.
"We are measured not by how many agents we train, but by how many humans trust them."
Who's behind this
A research lab is its people — and a discipline is only as strong as the track record of those who hold it.

Dmitriy Istomin
Co-founder, CEO. Product & operations. Serial entrepreneur with a successful exit; former CEO of Examus (AI proctoring, $2M ARR); Berkeley SkyDeck alumni.
LinkedIn

Taras Pustovoy
Co-founder. Trust framework architecture, personal agent methodology, vision. 25 years in EdTech and AI; co-founder of the National Platform of Open Education (2.7M learners); former Coursera content partner.
LinkedIn

Alexander Volkov, PhD
Co-founder, Research lead. HALA Framework, learning science, cybernetics.
LinkedIn

Mariam Mamedli, PhD
Co-founder. AI engineering, agent infrastructure. PhD (Sorbonne University & Paris School of Economics); 14+ years in AI and data.
LinkedIn
Gennady Koveshnikov
Founding Engineer. Builds the systems the research runs on — the pipelines, tooling, and instrumentation behind the measurements.
LinkedIn
Academic Council

Dale P. Johnson, PhD
Director, Digital Innovation — University Design Institute, Arizona State University.
LinkedIn

Irina Zingerman, PhD
Psychology & learning — process-oriented psychology. 15+ years in human–AI interaction design.
How we work
- Evidence classes on everything. Every claim on this site is tagged deployed & measured / prototyped / specified / frontier — including the claims about ourselves.
- Strict scales. By our own L0–L5 scale we score L0.9 today — a self-assessment. An easy scale would just tell everyone they're fine.
- Losing experiments in public. We published the experiment that disproved our own hypothesis. A lab that can't lose an experiment can't be trusted to win one.
The ecosystem
8Hats Lab — the research and measurement discipline. Agents University — where humans and agents learn it: our teaching arm, running on HALA (Human-AI Learning Architecture).
Contact
Press & partnerships: same address. Careers: not hiring publicly yet.
8Hats Lab Inc. · 131 Continental Dr, Suite 305, Newark, DE 19713-4324, United States · Lab: Singapore + Silicon Valley