$5M for VLC Creator's Robot Startup: Lightspeed Bets on Kyber

When a founder whose previous project reached 6 billion users raises money for something new, investors pay attention. Jean-Baptiste Kempf's pivot from VLC Media Player to Kyber represents a specific kind of founder bet that VCs covet: deep technical expertise from a proven open-source leader applied to a market—physical AI and robotics—that's still waiting for its infrastructure layer. Lightspeed's $5 million seed round isn't just capital; it's a thesis-level investment in the idea that real-time remote control will be as foundational to the robot economy as VLC was to digital video.

Jean-Baptiste Kempf, the architect behind VLC Media Player's 6-billion-download legacy, has secured $5 million in seed funding led by Lightspeed Venture Partners for his latest venture: Kyber, a Paris-based infrastructure startup built to solve one of physical AI's most fundamental bottlenecks—real-time remote control. The round, announced via a LinkedIn post from the American VC firm, positions Kyber at the intersection of video streaming expertise and the coming wave of autonomous machines that Kempf predicts will number in the "hundreds of millions" within years.

“

Lightspeed's $5 million seed round isn't just capital; it's a thesis-level investment in the idea that real-time remote control will be as foundational to the robot economy as VLC was to digital video.

Kyber's core product is an SDK that synchronizes video, audio, sensor data, and control inputs with minimal latency—an unglamorous but mission-critical piece of infrastructure for any system where the operator, the compute, and the physical action are geographically separated. The technology emerged from a side project Kempf built during his tenure as CTO at Shadow, the cloud gaming startup, where the challenge of streaming high-fidelity interactive content over unpredictable networks first crystallized into a broader thesis: that the same techniques enabling seamless cloud gameplay could underpin the next generation of robotics, drones, and autonomous systems.

The name Kyber—a nod to the lightsaber crystals in Star Wars—signals the company's obsession with speed. "If you control things in the real world, every millisecond matters," Kempf told TechCrunch, encapsulating the central engineering challenge. Unlike video streaming, where buffering is an annoyance, latency in physical control systems can mean the difference between a precise surgical maneuver and a catastrophic failure. Kyber's approach attacks this problem from two angles simultaneously: video-streaming optimization inherited from Kempf's VLC and Shadow experience, and IoT-level tuning that adapts performance to the available compute resources of each device at scale.

The timing of Kyber's emergence aligns with a broader industry shift toward what's being called physical AI—intelligent systems that perceive, reason, and act in the material world. Lightspeed, which has placed bets on foundational AI companies Anthropic and Mistral AI, clearly views Kyber as a complementary layer in this stack. The VC firm's public statement—"Physical AI is only as good as the underlying systems running it"—frames Kyber not as an AI company itself but as essential plumbing for an ecosystem that will include millions of robots, drones, and remotely operated vehicles.

But Kyber's addressable market extends well beyond AI-native applications. Kempf describes the platform's use case expansively: any scenario "where the person who's operating is not in the same place as the compute, which is not in the same place as the action." This encompasses telemedicine, industrial inspection, disaster response, agriculture, construction equipment, and defense—sectors where remote operation is already growing but where off-the-shelf middleware for ultra-low-latency control remains scarce. Large enterprises with sufficient resources—likely including the defense and logistics giants Kempf alludes to—have built proprietary versions of this software internally, validating both the need and the opportunity for a standardized, commercially available solution.