Axiomatic_AI

Axiomatic AI is a research-driven company developing AI systems designed to produce mathematically verifiable results rather than probabilistic guesses. Founded in 2024 by professors from MIT, the University of Toronto and ICFO Barcelona, and backed by Kleiner Perkins, the firm operates from Boston and Barcelona. Its stated aim is to deliver error-free outputs for scientific and engineering applications, where the reliability of a result matters as much as its speed.

The company's principal approach, Automated Interpretable Reasoning (AIR), combines reinforcement learning, large language models, world models and formal verification. By integrating formal proof systems with deep learning, AIR is intended to make AI systems prove their own correctness, replacing hallucinated outputs with results that can be checked mathematically.

Application areas include photonics, semiconductor design, hardware development and scientific research. The company frames its work around "Mission 10X30", a goal of reducing by tenfold the time engineers spend on technical challenges by 2030, and states a conviction that AI should support human reasoning rather than substitute for it. Engineering and research roles accordingly emphasise rigour, verification and reliability.