Antioch Raises $32M Series A Led by Greylock for Physical AI Verifier

On September 17, Antioch announced a $32 million Series A round led by Greylock, with participation from A*, Category Ventures, BoxGroup and Icehouse Ventures. Individual investors include Palantir CTO Shyam Sankar, Foxglove CEO Adrian Macneil, and NVIDIA executive and former Groq executive Ian Andrews.

Antioch is building what it calls a physical AI verifier, an environment that integrates customers' hardware, sensors, software, models, environments and operating constraints into a simulation and evaluation platform. The goal is to let developers predict whether a change will improve a physical system before running expensive real-world tests, enabling faster iteration cycles for robotics and other physical AI applications.

The company's approach combines classical simulation with learned world models. Classical simulation models known physical behavior from geometry, motion rules and hardware specifications, while learned world models are trained on real data to capture behaviors that are difficult to describe explicitly. Antioch believes end-to-end learned world models will ultimately prevail as more data becomes available.

Antioch was founded by a team with backgrounds at Tesla Autopilot, Meta Reality Labs and Google DeepMind. The company is already working with Amazon's Ring on intelligent perception, and Ring's chief product officer said Antioch's simulation results closely matched physical tests.