This industrial AI startup is winning over customers by saying it won’t get acquired

When industrial AI startup CVector meets with manufacturers, utility providers, and other prospective customers, the founders are often asked the same question: Will you still be here in six months? A year? 

It’s a valid concern in an environment where the biggest, richest tech companies are luring top talent with eye-watering salaries and increasingly targeting rising AI startups with elaborate acqui-hire deals. 

The answer that CVector founders Richard Zhang and Tyler Ruggles give every time is also the same: They’re not going anywhere. And that matters to their customers — a list that includes national gas utilities and a chemical manufacturer in California — which use CVector software to manage and improve their industrial operations.

“When we talk to some of these big players in a critical infrastructure, the first call, 10 minutes in, like 99% of the time we’re gonna get that question,” Zhang told TechCrunch. “And they want real assurances, right?”

This common concern is one reason why CVector worked with Schematic Ventures, which just led a $1.5 million pre-seed round for the startup. 

Zhang said he wanted to bring on investors that have a reputation for working on these kinds of hard problems in supply chain, manufacturing, and software infrastructure, which is exactly what Schematic is focused on as an early-stage fund. 

Julian Counihan, the Schematic partner who made the investment, told TechCrunch that there are a few ways startups can try to allay these kinds of concerns for customers. There are practical solutions — say, putting code in escrow, or offering a free, perpetual license to the software if an acquisition happens. But sometimes “it comes down to founders being mission-aligned with the company and clearly communicating that long-term commitment to customers,” he said.

It’s this commitment that seems to be helping CVector find early success.

Zhang and Ruggles each bring unique skills that play well with the type of work CVector provides its customers. One of Zhang’s earliest jobs was working as a software engineer for oil giant Shell, where he said he was often in the field “building iPad apps for people who’ve never used an iPad before.” 

Ruggles, who has a PhD in experimental particle physics, spent time working at the Large Hadron Collider “working with nanosecond data, trying to ensure very high uptime, being held accountable for downtime and rapidly troubleshooting.” 

“Those are places where you get to build up that kind of confidence, and that kind of background really helps give people some trust, some confidence in you,” Ruggles said.

CVector is more than just its founders’ résumés, though. The company has also been clever and resourceful since getting off the ground in late 2024. It built its industrial AI software architecture — what it refers to as a “brain and nervous system for industrial assets” — by leveraging everything from fintech solutions to real-time energy pricing data to open source software from the McLaren F1 racing team. 

They’re also taking different approaches on how to shape this brain and nervous system in real time with its customers. One example Zhang gave is with weather data. 

Changing weather conditions can have an impact on how high-precision manufacturing equipment works on a macro scale, but there are also knock-on effects to consider, he said. If it snows, that might mean the surrounding roads and parking lots get salted. If that salt gets carried into a factory on workers’ boots, it can have a tangible impact on the high-precision equipment that operators might not have previously noticed or been able to explain.

“Bringing those kinds of signals into your operations and your planning is incredibly valuable,” Ruggles said. “All of this is to help run these facilities more successfully, more profitably.”

CVector has already deployed its industrial AI agents in sectors like chemicals, automotive, and energy, and has its eyes set on what Zhang refers to as “large-scale critical infrastructure.” 

With energy providers specifically, Zhang said a common problem is that their grid dispatch systems are written in old coding languages like Cobra and Fortran that make real-time management challenging. CVector is able to create algorithms that can sit on top of those old systems and give operators better visibility into these systems with low latency.

CVector is small right now, with just an eight-person team distributed across Providence, Rhode Island, New York City, and Frankfurt, Germany. But they expect to grow now that the pre-seed is complete. Zhang did stress they’re recruiting only “mission-aligned people” who “actually want to make a career in physical infrastructure” — which will continue to make it easier to convince customers that the startup isn’t going anywhere.

While there’s a fairly straight line from what Zhang was doing at Shell to what CVector is up to now, it’s a bit more of a departure for Ruggles. But he said it’s been a challenge that he’s relished.

“I love the fact that instead of trying to write a paper, submit it, get it through the peer review process and get it published in a journal and hope that somebody looks at it, that I’m working with a client on something that’s in the ground and that we could be helping them keep it up and running,” he said. “You can make changes, build up features, and build new stuff for your customers — rapidly.”