Ex-Tesla founders raise $12.5M for Atomic's autonomous supply chain AI

Atomic raises $12.5 million Series A

Boston-based supply chain startup Atomic has closed a $12.5 million Series A round, lifting its total funding to just over $15 million, according to TechCrunch. Growth equity firm Klass Capital and Madrona Venture Group led the round. Jon McNeill, the former Tesla president whose DVx Ventures incubated Atomic, told TechCrunch that the company's annual recurring revenue has quintupled since the beginning of the year.

Atomic's software works out how much inventory a company should hold and where by simulating scenarios, then recommending or automatically choosing a response. Co-founders Michael Rossiter, the CEO, and Neal Suidan, the chief product officer, first built an early version of the system at Tesla during the 2018 Model 3 production ramp, when the automaker's spreadsheets could not keep up with how quickly planning had to change. Atomic has also brought on longtime Tesla planning director Jeff Goodrich as CTO and third co-founder.

From recommendations to autonomous purchasing

Since emerging from stealth last year, the product has evolved from an optimization platform that gives recommendations into one that makes decisions on its own, McNeill said in an interview with TechCrunch. The customer base has moved from pilots to what he called "DoorDash-scale customers," including DoorDash and HelloFresh. McNeill, who also sits on Atomic's board, said DoorDash now runs roughly 90% of its purchasing across hundreds of sites through the platform. For food-focused customers, the software's value shows up as less waste and spoilage.

Rossiter describes supply chain planning as a near-infinite optimization problem that keeps shifting under you. "AI can play the role of finding all of the best paths through that forest," he told TechCrunch.

Adaptability and fast onboarding drew investors

Two things attracted investors to the Series A, according to Rossiter: Atomic's ability to adapt a general model of supply chains across industries, and how quickly it can be deployed with new customers. Beyond food delivery, the company is working with consumer packaged goods firms and pushing deeper into mobility and manufacturing, which Rossiter described as a return to its Tesla roots.

McNeill said the board set a challenge of compressing onboarding until turning Atomic on would be "a non-event" for a customer. Suidan ran that effort and pushed the company's agentic AI to infer the "decision rules" a customer's staff followed even when they had never been written down anywhere. Once customers saw that the system understood their rules, McNeill said, they asked Atomic to simply make the decisions and free up their planners' time.

McNeill traced the underlying pitch back to Tesla, recalling that Elon Musk argued decision speed would separate the company from competitors such as Ford and Toyota because fast decisions compound, day over day, while rivals take weeks to make a first call.

Why it matters

Rossiter's stated ambition is to pull supply chain planning out of spreadsheets and into software that can actually make decisions. He notes that CFOs have driven this kind of modernization for finance data, while operating data rarely gets the same priority — partly because operational decision rules live in people's heads rather than in systems, which has historically made them hard to automate.

Atomic is also a concrete test of agentic AI moving from assistant to operator inside large enterprises. When a platform autonomously handles the bulk of purchasing at a company of DoorDash's scale, the open question is no longer whether AI can recommend supply chain actions, but how much of the decision itself organizations will hand over — and what oversight they will demand as that share keeps growing.