Applied Intuition

Applied Intuition sells a bet, not just a product: that software, not hardware, is what will finally make machines drive, fly, dig, and sail themselves. The company builds the digital plumbing (simulation, an operating system, and a full self-driving stack) that carmakers and defense agencies use to bring autonomy from a demo to the road.

Qasar Younis and Peter Ludwig founded it in 2017, one out of Y Combinator's COO seat and General Motors, the other out of Google Maps and Android Automotive. Eight years later it employs more than 1,000 engineers, counts 18 of the top 20 global automakers as customers, and closed a Series F in June 2025 that put its valuation at $15 billion. The pitch it likes to repeat is that most of autonomy's impact hasn't happened yet, and that Applied intends to own the infrastructure when it does.

Applied Intuition wants to be the AI backbone for anything with wheels, wings, tracks, or a hull. It groups its work into three layers: tools and infrastructure for building autonomy, an operating system for the vehicle itself, and a full self-driving system that handles perception, planning, and control.

The common thread is a closed-loop data flywheel. Real vehicles collect petabytes of sensor data, that data feeds simulation and model training, and the improved models push back out to the fleet over the air. Applied runs this across cars, haul trucks, mining rigs, tractors, drones, and military platforms, which is a much wider spread than most autonomy companies attempt.

Getting a car to drive itself once, on a closed track, is a solved problem. Doing it safely across millions of vehicles, in fog and snow and GPS-denied quarries, is where autonomy programs go to die. Applied's own engineering posts describe the failure mode in detail: fragmented tools, KPIs that arrive weeks after a test campaign ends, and rare edge cases (a very slow pedestrian, a moose on a dark road) that on-road testing almost never catches.

The other problem is hardware churn. Sensors and compute change on a cycle far shorter than a vehicle program, so an autonomy stack welded to one chip becomes a liability. Applied's answer is a hardware-agnostic platform plus simulation that turns real near-misses into thousands of parameterized test scenarios, so teams find the failure before the vehicle does.

The buyer is almost always an engineering or program leader inside a large, safety-critical operation. On the commercial side that means automotive OEMs and Tier 1 suppliers, plus trucking, mining, construction, and agriculture companies staring down a labor shortage (the average U.S. farmer is 58, and more than half the mining workforce is set to retire by 2029).

The other half of the customer base is national security. Through its defense arm and the EpiSci acquisition, Applied sells all-domain autonomy to the U.S. military and its allies, which is why many of its engineering roles require U.S. citizenship and a security clearance.

Applied's platform reads as an end-to-end stack rather than a pile of point tools. Tools for Vehicle Intelligence is where engineers build, simulate, and validate; Vehicle OS is the software layer that runs on the machine; and the Self-Driving System is the autonomy itself.

The newer pieces show where the company is pushing. Cabin Intelligence turns a tractor or haul-truck cab into something closer to a smartphone, and the Self-Driving System now ships domain-specific variants like SDS for Construction that keep running even when GPS drops out entirely.

Applied makes money the enterprise way: multi-year software and platform contracts with large OEMs, industrial operators, and defense agencies, not self-serve seats. There is no public price list, which is standard for a company whose deals are measured in vehicle programs and defense budgets.

Critics inside and outside the company note the flip side of that model. Its heavy customization work can make it feel, in the words of some reviews, closer to a consulting shop than a pure software vendor. The upside is stickiness: once an automaker builds a vehicle program on your stack, switching is enormously expensive.

Applied competes on breadth. Most rivals pick a lane (a simulation tool, a robotaxi stack, a single OEM partnership), while Applied sells across land, sea, air, and space and across commercial and defense at once. That diversification is its main hedge and its main story.

The advantages it leans on are real ones: a hardware-agnostic platform that survives sensor and silicon changes, deep entrenchment with 18 of the top 20 automakers, and a defense position sharpened by acquiring EpiSci, whose software flew the Air Force's X-62A VISTA in an autonomous dogfight against an F-16.

The proof is in who's building on the stack. Toyota's Woven Planet Group uses Applied's simulation to test its automated driving system across tens of thousands of scenarios in the cloud; Komatsu is developing its next-generation autonomous haulage system with Applied; and Heidelberg Materials is running SDS-powered autonomous trucks in an Australian quarry where GPS routinely drops out.

At the company level the scale speaks too: more than 1,000 engineers, 18 of the top 20 global automakers as customers, and a global fleet collecting petabytes of real-world driving data. Case studies also name AISIN, Isuzu, May Mobility, and Caterpillar.

The praise from employees is consistent and specific: genuinely hard, interesting work, unusually high talent density, and a fast culture where ideas turn into shipped work quickly. People who like it really like the mission and the caliber of colleagues.

The criticism is just as consistent. Reviews describe 50 to 60 hour weeks, weekend crunch tied to customer deadlines, and a work-life balance rating that sits well below the overall score. Others flag below-market pay, thin equity upside after later rounds, and high turnover, with several reviewers saying the average tenure runs under two years. It reads like a place that pays in impact and learning more than in comfort.

Applied's valuation more than doubled in a year. A $250M Series E in March 2024 pegged it at $6 billion; a $600M Series F in June 2025, led by BlackRock and Kleiner Perkins, took it to $15 billion.

The backer list is a who's-who: Andreessen Horowitz, General Catalyst, Kleiner Perkins, BlackRock, plus sovereign funds like the Qatar Investment Authority and Abu Dhabi Investment Council. The Series F framing was explicit about where the money goes, which is deeper into defense, and the EpiSci and Reblika acquisitions in 2025 followed the same line.

Applied's story is one of the most ambitious in autonomy: own the software layer for every moving machine, commercial and military, on every continent. The Series F, the OEM roster, and the defense acquisitions all point the same direction, and the labor shortages in mining, farming, and construction give the pitch real economic gravity.

The open questions are the ones its own reviews raise. Can a company selling heavily customized platforms scale like software rather than consulting, and can it hold onto the engineers it's burning through fast? If Applied answers both, a $15 billion valuation starts to look like a floor. If it doesn't, breadth becomes the thing that stretches it.

Applied wears its talent density as a badge: more than 1,000 engineers, 40-plus former CTOs, and 30-plus former founders on staff. The internal value it names first is speed, and the culture reflects it, high agency and fast-moving, with engineers owning hard problems early.

The trade-offs are baked in. This is an in-office company that expects people at their desks five days a week, with global hubs from Sunnyvale to Munich, Bangalore, Seoul, and Tokyo. Defense work adds citizenship and clearance requirements to a chunk of the engineering roles. The benefits are strong on paper, and the JDs are candid that the job runs at startup pace.

Applied discloses base ranges on most roles and pairs them with equity and a full benefits package. Engineering bases in the U.S. run wide, roughly $80K at the low end to the mid-$400Ks for senior and specialized roles, with product, design, sales, and operations clustering lower. UK and EU roles are posted in local currency.

Every eligible role states that total comp includes equity in the form of options or RSUs on top of base, plus a 401(k) match. Worth knowing from employee reviews: some feel base sits below market and that later-stage equity carries less upside than it once did, so weigh the whole package, not the headline number.

Applied Intuition, Inc. is the registered entity, headquartered in Sunnyvale, California, and it operates as a federal contractor, which shapes a lot of its hiring. Many defense-facing engineering roles require U.S. citizenship and an active or eligible Secret clearance.

On data, its privacy notice states plainly that it does not sell, rent, license, or lease personal information, and it names Google Analytics among the third-party services it uses. The company's engineering work also leans hard on automotive safety standards like ISO 26262, which shows up across both its product messaging and its job requirements.

The through-line in Applied's 2025 and 2026 coverage is expansion: a $600M Series F at a $15 billion valuation, the acquisition of defense-autonomy firm EpiSci (whose software flew an Air Force jet in an autonomous dogfight), and a steady drumbeat of industrial deployments with Komatsu, Heidelberg Materials, and Toyota's Woven Planet.

The company also stood up a UK subsidiary with a planned investment and keeps publishing detailed engineering posts on how its platform actually works. Links below run from press coverage to its own technical writing.