Home›Companies›Axiomat
Axiomat

AI that never hallucinates.

Axiomat makes it impossible for AI agents to make mistakes when querying a company’s data. It formally verifies that the agents’ queries are correct before employees or customers rely on the answers. It is becoming increasingly important for companies to rely on AI without the need for humans to manually check the answers. This has created a need for systems which can guarantee correctness without limiting the agent’s capabilities. Axiomat provides the infrastructure to prove that answers are correct, starting with data queries and analysis. It brings mathematical guarantees to the agents working with your data. The team brings direct experience solving AI hallucinations through Ethan’s work as an Applied Scientist at AWS, where he worked on formally verifying AI responses. They also bring research experience in AI, LLMs, and formal methods from EPFL, the University of Toronto, and Georgia Tech.
Active Founders
Ethan Kharitonov
Ethan Kharitonov
Founder/CEO
Ethan is a founder of Axiomat (YC F26). Previously, he was an Applied Scientist at AWS researching formal methods and AI verification - the only employee in Amazon’s history to move from a software role into applied science within eight months of joining. He has published research in formal methods and left his PhD at Georgia Tech to build Axiomat. He holds a B.S. in Mathematics and Computer Science from the University of Toronto, where he also led research in functional analysis and topology.
Jack Naimer
Jack Naimer
Founder/COO
Jack is a founder of Axiomat (YC F26). Previously he was a Master's Research Scholar in AI at EPFL, where he led pretraining of a large-scale DNA foundation model for drug discovery and cellular engineering. Before that he built the first DNA reasoning model in Bo Wang's lab, and he led the development of a machine learning model for the Peter Munk Cardiac Centre. He holds a B.A.Sc. in Engineering Science, majoring in Machine Intelligence, from the University of Toronto.
Elad Perlman
Elad Perlman
Founder/CTO
Elad is a founder of Axiomat (YC F26). Previously, he worked as a full-stack engineer at Survalent, developing real-time control software for utilities. He holds a B.S. in Mathematics and Computer Science from the University of Toronto, where he graduated with high distinction.
Company Launches
Axiomat – Agents that cannot hallucinate
See original launch post

Hey YC 👋! We’re Ethan, Jack and Elad, founders of Axiomat.

TL;DR

We’re building formally verified AI that cannot hallucinate. Ever. We’re starting with queries over structured data. Axiomat lets your customers and employees ask AI about their data, even when a wrong answer is unacceptable.

https://youtu.be/GkoRgSpmd_c

The problem

AI can make mistakes

AI is ridiculously smart, but even world-class experts can make mistakes. When a wrong answer is unacceptable, the possibility of error forces you to check your agent’s work, restrict what your agents can do, or, in the worst cases, keep AI out altogether.

This tiny disclaimer does a lot of damage:

People want to use AI anyway

Dashboards are dead. Your users don’t want to learn your interface, and their needs won’t always fit the pages you’ve built. Agents let them ask for what they need without waiting for your team to build another screen. But if the answers “might be wrong” and need to be double-checked, they’re back to the work the agent was supposed to handle.

The solution

Axiomat gives AI agents a mathematical guardrail against all hallucination. We use your business rules to turn English questions into mathematical definitions, then formally verify in Lean that the queries agents make match those definitions. If we can’t prove it, the query doesn’t run. The agent retries until it has a proven answer or tells you it couldn’t find one.

Before it answers, Axiomat shows you how it understood your question. It translates the mathematical definition back into unambiguous English, so you see the exact interpretation the proof covers (and this translation is done with code, not AI!). You never have to wonder whether Axiomat answered a different question than the one you asked.

You can use Axiomat to launch an assistant you’ve held back or expand one you’ve had to restrict. Those same verified queries can supply the information an agent needs to manage finances or run a business, making this a first step toward trusting AI with those larger tasks.

Imagine what we could do with AI if we could actually rely on it.

The ask

If correctness is keeping your AI product unreleased or your agent restricted, email founders@axiomat.com to discuss using Axiomat for the data-querying use case you can’t safely offer today.

YC Photos
Axiomat
Founded:2026
Batch:Fall 2026
Team Size:3
Status:
Active
Location:San Francisco
Primary Partner:Nicolas Dessaigne