A former Tesla Optimus engineer has resolved a trade secret lawsuit and secured $11 million to develop robotic hands.

      TL;DR: Jay Li, former lead for Tesla's Optimus project, has resolved a trade secret lawsuit with Tesla and secured $11 million in funding for his startup, Proception, which specializes in dexterous robotic hands.

      Proception, a robotics startup established by ex-Tesla Optimus engineer Jay Li, has settled a year-long trade secret lawsuit with Tesla and raised an $11 million seed funding round led by First Round Capital to develop dexterous robot hands. The startup announced to TechCrunch that it is now delivering the initial batch of its high-dexterity hands to researchers and robotics companies, while also accepting more orders. The round saw participation from Y Combinator and early-stage fund BoxGroup.

      In June 2025, Tesla filed a federal lawsuit against Li and Proception in Northern California, claiming that Li had downloaded confidential files related to robotic hand actuation onto personal devices prior to resigning and launching the startup just six days later. The lawsuit pointed out that Proception’s hands resembled Tesla’s internal designs. After several months of legal battles, the parties reached a settlement, leading Tesla to dismiss the case earlier this month.

      Li shared with TechCrunch that he views the ordeal as a “test of resilience” and believes that the company has come out stronger as a result. He expressed that he wouldn’t be shocked if Tesla eventually seeks assistance from Proception regarding their own robotic hand issues. Tesla has not commented on the situation.

      The challenge of dexterous manipulation—grasping, rotating, and manipulating objects with precision akin to humans—remains one of the most difficult issues in robotics. Even Elon Musk has described robotic hands as one of the biggest engineering hurdles yet to overcome. Kevin Lynch, director of Northwestern University’s Center for Robotics and Biosystems, indicated to the Wall Street Journal last year that his team estimates it will take another decade before robot hands are functional and practical enough to perform human-like tasks.

      Li believes Proception can accelerate this progress due to its innovative method of gathering training data. Typically, companies training humanoid robots use teleoperators, where a human with a virtual reality headset remotely operates a robot, and the system learns from these actions. However, Li points out that this method lacks tactile feedback from the objects interacted with and is also limited by the number of available robots.

      Proception’s approach involves a sensor-equipped glove that captures human hand interaction data without requiring a robot. This glove additionally serves as the sensor-laden "skin" for the robotic hand Proception is developing, which boasts 22 degrees of freedom and multiple joints per finger. Li contends that this combination of scalable data collection and advanced hardware is what the market needs.

      The market for dexterous hands has seen substantial investment this year. China's Linkerbot, which dominates 80 percent of the global market for high-degree-of-freedom hands, is aiming for a six billion dollar valuation after shipping over 1,000 units monthly. Meanwhile, Genesis AI, a European startup, raised $105 million for its wheeled robot with dexterous hands, and Chinese companies like Xynova have garnered nearly one billion yuan in funding.

      Proception anticipates that many humanoid robot manufacturers will prefer to purchase hands rather than produce them internally, similar to the automotive industry's approach to specialized components. First Round partner Bill Trenchard, who led the investment, remarked to TechCrunch that dexterous manipulation is “the last mile in achieving true robot performance.” He also commended Li’s leadership during the challenging lawsuit with Tesla.

      Tesla has explored manufacturing Optimus at its Shanghai Gigafactory and has deployed over 1,000 Gen 3 units across its facilities, though the hands remain a significant limitation. Musk has set a target price between $20,000 and $30,000 per unit and projected production scaling to tens of thousands by 2028. Whether Tesla will produce its hands in-house or source them from companies like Proception is an ongoing question in the humanoid robot supply chain.

      Currently, more than 150 companies are vying for the humanoid robot market, with billion-dollar valuations becoming commonplace and only 23 percent of enterprise buyers satisfied with the products on the market. In this context, a startup focusing on a crucial component that everyone agrees is challenging to perfect has a strong pitch, even at the seed stage. The investment made by First Round Capital hinges on whether Proception can evolve from its initial shipments to a position where it influences how a broad category of machines utilize their hands.