Applied Intuition

Applied Intuition develops software and AI for moving machines, including cars, trucks, mining and construction equipment, agricultural machinery, and defense systems. Its technology spans simulation and development infrastructure, vehicle operating systems, and autonomy models. With the July 2026 introduction of Dana, it brought agents into the workflows used to develop and validate those systems.[1][10]

That breadth followed an earlier focus on tools for autonomy developers. The company's beginnings explain the simulation and data infrastructure underlying its later products.[1][10]

Qasar Younis and Peter Ludwig founded Applied Intuition in 2017, serving as CEO and CTO respectively. Younis had worked as an automotive engineer at General Motors and Bosch, founded a company acquired by Google, and served as COO of Y Combinator. Ludwig's Google work included Google Maps and Android Automotive.[3][2]

In their April 2026 interview, Ludwig recalled that the first customers were robotaxi companies and that the initial work centered on simulation and data infrastructure. Younis described tooling as a way to build and sell technology to customers who wanted to develop autonomy themselves. By that interview, Ludwig said the portfolio had grown to more than thirty products.[1]

Applied Intuition completed its acquisition of Embark Technology on August 2, 2023, in an all-cash transaction with an equity value of approximately $71 million. The May agreement announcement outlined plans to incorporate Embark's internal tools, data, and software assets into products for trucking and automotive customers; Embark planned to retire its test fleet as part of the transaction.[6]

On February 6, 2025, Applied Intuition announced its acquisition of EpiSci, expanding its defense autonomy capabilities across land, air, sea, and space. EpiSci became a wholly owned subsidiary, and co-founder and CEO Bo Ryu, co-founder Tamal Bose, and its team joined Applied Intuition. The planned integration paired EpiSci's tactical AI with Applied Intuition's simulation, validation, and data-management software. EpiSci brought prior work in AI-piloted fighter aircraft, drone-swarm testing, and maritime demonstrations.[7]

Released on January 6, 2025, Neural Sim addressed a limitation of recorded sensor replay: when a simulated vehicle substantially departs from the recorded position and orientation, the original sensor data no longer represents what it would see. Neural Sim reconstructs virtual worlds from drive logs and renders camera, lidar, and other sensor data as the tested system changes its pose. This enables closed-loop testing, in which the driving system's actions affect its subsequent inputs.[8]

Ludwig explained that simulation still requires repeated calibration against real-world results. His example was actuator heating: a reinforcement-learning policy can account for a motor's temperature only if the simulator represents it; otherwise, apparently successful simulated motions may overheat the physical machine. The founders described real-world testing as a continuing part of development.[1]

Vehicle OS extends below the cabin interface into scheduling, memory management, middleware, message passing, and reliable networking. Ludwig described controlling sensor and actuator latency, providing fail-safes, and enabling reliable updates to critical vehicle software. He recalled that dissatisfaction with available systems prompted Applied Intuition to build its own operating-system technology. Customers can license the operating system while using another supplier's autonomy technology, or combine it with Applied Intuition's autonomy stack.[1]

Applied Intuition announced SDS for Automotive on August 19, 2025. The learned end-to-end ADAS platform integrates perception, planning, and control while exposing intermediate outputs, including detected objects and scene understanding, for inspection and traceability. It integrates with Vehicle OS and gives manufacturers full source-code access for customization and validation. The announcement promised a complete L2++ feature set and a pathway to L3 and L4; those were capability plans, rather than evidence of deployed L4 passenger cars.[9]

The company's research and engineering work includes world models, autonomy models, and neural simulation. In April 2026, Ludwig described neural simulation as combining Gaussian splatting and diffusion methods, with speed and cost determining whether reinforcement learning is practical. He distinguished large offboard models from smaller onboard derivatives constrained by millisecond response times, power, and embedded compute.[1]

On June 17, 2025, Applied Intuition announced that it had closed its Series F at a $15 billion valuation, with BlackRock and Kleiner Perkins leading the round and a tender offer accompanying the financing. Axios, citing an unnamed source, reported $200 million in primary capital and $400 million in secondary transactions.[4][5]

The financing announcement also described a strategic partnership with OpenAI to combine its large language models with Applied Intuition's vehicle-intelligence technology. Proposed applications included natural, context-aware driver interactions, trip planning, and personalized infotainment.[4]

In February 2025, Applied Intuition reported that eighteen of the top twenty global automakers trusted its solutions. In the April 2026 interview, Younis confirmed approximately one thousand engineers and said engineering accounted for 83 percent of the company.[7][1]

Dana connects sensor-data curation, workflow orchestration, simulation, evaluation, and analysis through agents and a common interface. Developers can run open-loop replay, reconstruct scenes for closed-loop neural simulation, generate adversarial behaviors with reinforcement learning, and vary weather or lighting with world models. Results retain workflow lineage so engineers can trace failures and feed findings into the next development cycle.[10]

At the July 21, 2026 launch, Applied Intuition reported internal Dana use since 2025 and limited early access for customers including Isuzu Motors and Komatsu. It said it had rebuilt a significant portion of its autonomy-tools platform's core functionality in six months to make it available with agentic capabilities. Its reported twentyfold acceleration in development cycles described internal experience.[10]