Lyte, a cutting-edge startup focused on robotics and artificial intelligence, has successfully secured $165 million in funding, boosting its valuation to $1.6 billion—an impressive increase that has nearly tripled its worth in less than a year. This latest investment elevates the company's total funding to $272 million.
Based in Mountain View, California, Lyte finds itself in the company of a rising number of startups attracting major investments aimed at creating the hardware and infrastructure necessary to integrate AI into the physical realm.
The company is backed by a combination of experienced semiconductor entrepreneurs and engineers. Avigdor Willenz, one of Israel's notable figures in the semiconductor industry, serves as the founding investor and chairman. The team also includes engineers who played pivotal roles in developing the depth-sensing technologies underpinning Apple's Face ID.
Maverick Silicon co-led this funding round, which also saw participation from Fidelity Management & Research, Atreides Management, Key1 Capital, and Ora Global. Exor Ventures, an early supporter, continues to invest in Lyte.
This new funding comes just eight months after Lyte announced its presence to the public on January 5 with an initial $107 million in funding. Since that announcement, the startup has more than doubled its investments and surpassed the $1 billion valuation mark.
Willenz’s partnership reinforces Lyte's connection to a legacy of influential Israeli semiconductor firms that have become significant players in technology innovation. He established Galileo Technologies, which specialized in networking chips and went public on Nasdaq before being acquired by Marvell in a deal worth approximately $2.7 billion. Willenz was also an early investor in Annapurna Labs, which created processors for Amazon’s cloud services, and he co-founded Habana Labs, an Israeli company specializing in AI processors, which Intel acquired in 2019 for roughly $2 billion.
Lyte's strategic focus is on a new layer of technology poised to grow in relevance as artificial intelligence expands beyond data centers. "Lyte is building at the right layer, at the right moment,” Willenz remarked upon the company’s debut. “I’ve witnessed how fundamental technologies can revolutionize entire industries. What is particularly noteworthy here is the depth of expertise in the team and their commitment to addressing perception as a system, where substantial value is generated."
The founders of Lyte each bring significant experience in perception technology to the table. CEO Alexander Shpunt co-founded PrimeSense, the Israeli company behind the technology that powered Microsoft's Kinect. Apple acquired PrimeSense in 2013 for $350 million, after which its technology was integrated into Apple’s depth-sensing platform. Following this, Shpunt served at Apple as a Distinguished Engineer, focusing on perception systems utilized across various products.
CTO Arman Hajati's journey diverged from Shpunt’s but is equally distinguished. Hailing from Iran and educated at the University of Tehran, he honed his engineering skills in the U.S. and eventually became a senior engineering manager at Apple, specializing in integrated sensing technologies that inform Lyte’s current methodology.
The third key founder, Yuval Gerson, is another PrimeSense veteran who transitioned to Apple as a MEMS system architect before taking on his role at Lyte as an engineering fellow and vice president.
While Lyte is not developing a physical robot, its aim is to create foundational technology that enables robots and autonomous machinery to better perceive and interact with their environments. Lyte’s premier offering, LyteVision, merges 4D sensing, RGB imaging, and motion awareness into a comprehensive perception platform. The company is also focused on crafting custom silicon and software to seamlessly integrate these elements.
Lyte is tackling a critical challenge as robotics continue to evolve. Although AI models have made strides in processing language, images, and other digital data, autonomous systems must also navigate and interpret ever-changing physical surroundings. Typically, robotics companies gather various sensors and components from multiple suppliers but often face the daunting task of integrating and calibrating these elements while developing software to process the data effectively.
