Axiom Math

Overview

Axiom Math is an AI company focused on formal mathematics and formal verification, founded by Carina Hong and based in downtown Palo Alto. The company was approximately 8 months old at the time of UNLOCK 2026. The founding team draws from math olympiad backgrounds and includes PhD mathematicians. The core product, Axiom Prover, achieved 100% on the Putnam exam and 99% on the Verina code verification benchmark. The system autonomously proved that the Ramanujan tau function misses 100% of prime numbers — a result requiring deep number-theoretic reasoning.

A key operational demonstration: Axiom Math receives an open mathematics problem at 10am and returns a fully verified Lean proof by 2pm. The company identified a structural insight with major commercial implications: the mathematical structure underlying formal proof and code verification is deeply shared, meaning a system trained on mathematics transfers to chip and code verification with no additional tuning. This positions Axiom Math for the EDA (electronic design automation) market, where chip verification cycles are currently 3–4x longer than design cycles and verification teams are 3–4x larger than design teams.