Maven Robotics emerges from stealth with $100 million to scale its industrial robots

Maven Robotics, a robotics startup founded in 2024 by brothers Hamza and Khalid Derbas, has announced it is coming out of stealth mode after closing a $100 million round with participation from RoboStrategy, LocalGlobe, Vine Ventures and XTX Ventures. The funds will go toward building 250 units of its third-generation robot and starting work on the design of a fourth-generation platform.

As CEO Hamza Derbas tells TechCrunch, the company started from practically nothing: "a drawing of a robot and a team of people." They landed their first major customer in an unconventional way: upon learning that a large consumer goods company with logistics needs was meeting with four competitors in the robotics sector, Derbas managed to secure a meeting and, instead of talking about robots, asked to visit their factories and warehouses. That direct observation of the workflows allowed them to identify where they could add value immediately, which ultimately convinced the customer to choose them over rivals that already had robots up and running.

After two years working with that customer and other partners, Maven says it currently has up to eight robots operating 16 hours a day with uptime of 99% or better. The company's central thesis is not to solve the problem of an isolated robot, but to automate the process end to end: the system connects on one side to the warehouse management software (WMS) and on the other delivers the product directly into the trucks.

On the hardware side, Maven's robots move on wheeled bases at up to 10 miles per hour (about 16 km/h) and have two arms capable of lifting up to 30 kilograms. Their main task is so-called "mixed palletizing": wooden pallets loaded with goods arrive from different factories at a distribution center, and the robot builds a new pallet combining different products according to what each store needs. That combination is adjusted very quickly, since retailers want to be able to change the product mix based on real-time demand barely 48 hours after restocking the shelves — a task that today is carried out almost exclusively by human labor.

The article notes that Derbas comes from a career in automotive engineering focused on electric vehicles and spent nine years in Apple's special projects group, widely believed to have been responsible for developing a self-driving car before it was disbanded in 2024, although he avoids confirming specific details about that work. His brother Khalid, with prior experience in private equity, serves as the company's chief financial officer. As is the case with other "physical AI" firms, Maven leans on veterans of autonomous driving, a sector that has developed sophisticated methods for training autonomous hardware from real-world data, using pipelines that make it possible to retrain models, evaluate results and redeploy improvements in cycles of minutes or hours.

RoboStrategy investor Jack Pearson argues that what sets Maven apart from other companies in the sector is its origins in real industrial systems, rather than a research-oriented culture or one centered on a specific technology architecture. The article cites as a comparison Agility Robotics, a company that will go public this fall through a SPAC merger valued at $2.5 billion and that likewise focuses on safety and specific industrial workflows. Derbas, however, draws a clear distinction: Agility's robots walk on two legs, a design that, in his view — despite saying he respects the company — "makes no sense" for the tasks they tackle, being "very complex, unreliable and adding unnecessary cost," stressing that in this business "return on investment is what rules."

Looking ahead, Maven acknowledges that, while palletizing represents an estimated $80 billion market, its next target — materials handling and the automation of manufacturing tasks — will require robotic capabilities that do not yet exist today. To get there, the company will combine its own systems with outside suppliers and has even developed gripper-style gloves that let human workers mimic the kind of grasp they are looking to replicate in their future end effectors. Despite describing itself as a maker of general-purpose robots, its strategy is to advance problem by problem: according to Derbas, every industrial problem of sufficient size constitutes a multibillion-dollar market in its own right, which generates enough data to master each skill incrementally.

The piece concludes by noting that this gradual approach, focused on solving one customer problem at a time, could be the most viable path to introducing robots into the real world of work, though it also carries the risk of being suddenly overtaken by the next powerful "physical AI" model to emerge from the frontier labs. Derbas settles this tension with a line that sums up the company's philosophy: "We're not in the model race: we're in the race to solve industrial work and make this work at the scale the world needs."