Its initial focus is mixed palletizing, a warehouse process in which products arriving from different factories are reorganized into new pallets based on what individual stores need. Today, much of that work is still performed manually.
“Within 48 hours of them putting the stuff on the shelves, they want to change the mix based on real-time demand,” CEO and co-founder Hamza Derbas said. “It’s all done with human labor today, running around the warehouse picking one of this, one of that.”
Maven began pursuing that type of work before it had a finished robot. In 2024, the startup secured an early customer after Derbas persuaded a large consumer goods company to let the team study its factories and warehouses.
Rather than building around a single predefined robotics problem, Maven designed its system around completing broader workflows. Its robots can connect with warehouse management software on one end of the process and prepare products for shipment on the other.
“We’re not trying to solve a single robot problem,” Derbas said. “We’re trying to autonomously take on the task end to end.”
Derbas previously worked in automotive engineering and electric vehicles before spending nine years at Apple. He later founded Maven with his brother Khalid Derbas, who serves as chief financial officer.
The startup also borrows techniques developed in autonomous vehicle programs. Data from robots already working in customer facilities can be returned quickly, used to retrain models, evaluated and then redeployed as part of a continuous improvement cycle.
That operating approach is central to how Maven is positioning itself among a growing number of robotics companies. Investor Jack Pearson of RoboStrategy said the company’s strength comes from experience with industrial systems rather than a research-first culture built around a particular architecture.
Maven has also made deliberate hardware choices around reliability and cost. Derbas argues that wheeled robots are better suited to the company’s current industrial tasks than two-legged machines, which he described as more complex and expensive.
“ROI is the name of the game here,” he said.
The company now wants to move beyond palletizing. Its next phase centers on collecting more data and developing manipulation capabilities that could allow its robots to handle a wider range of materials before eventually moving into automation and fabrication tasks.
To help train those systems, Maven plans to combine data from its own robots with third-party sources. It has also created pincer-like gloves that allow people to perform movements resembling the actions expected from future robotic grippers.
Despite describing its machines as general-purpose robots, Maven is taking an incremental route toward that goal. Rather than attempting to solve many industrial tasks at once, the company plans to add capabilities one workflow at a time.
“We’re grounded in solving one customer problem at a time,” Derbas said. “If you focus on solving problems and you pick sizeable problems, each problem is a multi-billion-dollar market.”
For now, that means proving the machines can perform repetitive warehouse work reliably before Maven pushes into more technically demanding forms of industrial automation.
“We’re not in the race for models—we’re in the race to solve industrial labor and make this work possible at the scale the world needs,” Derbas said.
This analysis is based on reporting from TechCrunch.
Image courtesy of Maven.
This article was generated with AI assistance and reviewed for accuracy and quality.
