Forbes:

Axiom Math, an AI startup founded by former Stanford PhD student Carina Hong, aims to develop an "AI mathematician" capable of solving complex problems, generating rigorous proofs, and independently verifying its solutions. Hong conceived the idea during a conversation with Shubho Sengupta, formerly an AI researcher at Meta. The company envisions its AI ultimately proposing and proving new mathematical conjectures, thereby contributing to novel knowledge creation. Hong believes that "math is the perfect sandbox for building superintelligence," according to Forbes. Axiom Math is converting mathematical content from sources like textbooks, academic papers, and journals into a software program, enabling the AI to both formulate new problems and subject its solutions to formal testing and verification. The company, currently valued at $300 million, has secured $64 million in seed funding, led by B Capital, with participation from Greycroft, Madrona, and Menlo Ventures. In its first year, Axiom Math has assembled a team of 10 full-time employees, including several former researchers from Meta’s Fundamental AI Research (FAIR) lab. Notable additions include Francois Charton, formerly a mathematics-focused large language model researcher at Meta, known for solving a century-old mathematical problem, as well as Aram Markosyan, an AI research scientist who previously led safety and fairness research at Meta, and Hugh Leather.