Overview
Salaries
Products
People
Growth
Offices
Financials

Overview

Autodidakt is building reinforcement learning systems for faster, high-fidelity physics simulation at the solver level.