As artificial intelligence becomes more widely used in science and engineering, a key problem remains: most AI systems can generate ideas or designs, but they cannot prove that those ideas obey the laws of physics.
A new startup is trying to solve that challenge. Axiomatic AI, a startup focused on building verified AI for science and engineering, has raised $18 million in a seed funding round, bringing its total funding to $25 million.
The round was led by Engine Ventures, with participation from Kleiner Perkins, Big Sur Ventures, G Vision Capital, Propagator Ventures and Liquid 2 Ventures.
The company plans to use the new capital to expand deployments with enterprise customers and further develop its platform for complex scientific and engineering workflows.
Building AI that can verify engineering decisions
Axiomatic AI was founded in 2024 by a group of well-known researchers in physics and engineering. The founding team includes Dirk Englund, Frank Koppens, Joyce Poon and Marin Soljačić.
Led by Jake Taylor, Axiomatic AI is developing a platform designed specifically for technical fields such as semiconductors, photonics and advanced manufacturing, where systems must follow strict physical laws.
Taylor previously served as Assistant Director for Quantum Information Science at the White House and later worked as a senior advisor at the National Institute of Standards and Technology.
The company’s core technology, called Axiomatic Intelligence, combines advanced AI models with mathematical and physics-based verification methods. The goal is to create AI systems that not only generate ideas but also prove their reasoning.
Traditional generative AI tools can produce answers that appear convincing, but they cannot always verify whether the output is physically correct or logically consistent. Axiomatic’s platform aims to solve this by checking results against physical principles and engineering constraints.
This allows engineers to automate complex workflows while still verifying accuracy at several levels, including fundamental physics, design rules and logical reasoning.
“We are defining the standard that science and engineering AI must meet,” said Jake Taylor, CEO of Axiomatic AI. “As demand for the hardware underpinning our economy accelerates, machine learning systems must move beyond assistance into accountable collaboration. AI that cannot justify its reasoning to the level needed for engineering cannot scale into high-stakes technical domains. Our focus must be on shifting the baseline of technical intelligence to verifiable outcomes that connect to physical reality.”
Early users include major industry players
Axiomatic AI is already working with several large organisations through an early access program. These include companies involved in semiconductor manufacturing, chip design, and photonics technologies, along with collaborations with major research institutions.
The company’s long-term goal is to become a central platform used by organisations that build complex hardware and scientific systems.
“Science and engineering are the backbone of modern civilisation. The shift from prediction to provable reasoning will define the next era of AI adoption in critical industries,” said Israel Ruiz, President and General Partner at Engine Ventures. “Axiomatic is building the infrastructure layer that makes that shift possible.”
