Last month, a faulty CrowdStrike update took down airports, 911 call centers, and hospitals, highlighting how a faulty update can affect critical infrastructure. If the update was for something like a self-driving car or warehouse robot, the impact of a faulty update could be even more severe.
Early-stage startup Trace Machina is trying to avoid such scenarios with advanced simulation software that lets developers test updates in a more realistic simulated environment. The company emerged from stealth on Thursday to announce a $4.7 million seed round investment and an open-source tool called NativeLink.
CEO and co-founder Marcus Egan says the company is developing a native Rust-based system that will help test and validate software for autonomous systems, like self-driving cars and automated warehouse equipment, before deploying them in real-world environments.
“The way we solve this problem is to provide a native link between the developer and the autonomous vision,” Egan told TechCrunch, which is exactly why the company's first product is called NativeLink.
“As developers move from building web apps to building robots, it's clear that existing development toolkits like Docker, Kubernetes, etc. are not enough. Engineers need to be able to experiment and test directly on their local hardware,” he said.
“NativeLink fills that gap, providing engineers with a staging environment where they can run simulations in resource-constrained environments such as hard-to-source embedded Nvidia GPU chips for robots, self-driving cars and edge devices.”
According to Egan, in the past, companies had to build these environments themselves, and they were limited to well-funded self-driving car manufacturers and hyperscalers like Google. He wanted to build a system that was as close to the hardware as possible, or “near metal,” and make it available to any company.
“There's a lot of people who've been down this path, but nobody has had direct access to the hardware to run it. And frankly, these virtualization layers, these abstraction layers have always been there, which have made it easier for companies to build and iterate on these systems. We've just had to pay the price of being closer to the hardware,” he said.
Egan's background includes working at MongoDB, where he helped develop the company's first AI product, Atlas Vector Search, while co-founder Nathan Brewer worked at Google X, the company's experimental moonshot project center, and at Toyota Research Institute, where he worked on self-driving cars.
Egan, who is black, has had to deal with racism throughout his career, but remains focused on growing his company. “I've had to deal with racism, but I don't care. I'm focused on my goals. No one can stop me or tell me how things should go. From that perspective, I'm very grateful for that, because a lot of people like me don't have that freedom,” Egan said.
He's had to overcome other obstacles in his life besides racism: As a teenager, he was in a terrible car accident that left him seriously injured and unable to walk or talk, but he recovered, went to college to become an engineer, and eventually launched this startup.
The $4.7 million seed round was led by Wellington Management, with participation from Samsung Next, Sequoia Capital Scout Fund, Green Bay Ventures, Bellissimo Ventures, and several prominent industry angel investors.