Kyber Raises $5 Million to Scale Low-Latency SDK for Real-Time Control of Robots and Drones
Kyber has raised $5M to scale a low-latency SDK for real-time control of robots, drones and remote IT, pairing open-source software with enterprise tools.
Jean-Baptiste Kempf’s Paris-based startup Kyber announced a $5 million funding round led by Lightspeed to accelerate development of a low-latency software stack designed to control remote devices in real time. The company’s platform synchronizes video, audio, sensor feeds and operator inputs to reduce lag when humans or AI need instant control, and Kyber’s open-source roots remain central to its go-to-market strategy. Investors say the move signals growing demand for infrastructure that supports “physical AI” at scale.
Lightspeed Leads Strategic $5 Million Round
Kyber’s latest financing was anchored by Lightspeed, the venture firm that has backed several high-profile AI companies. The capital will be used to expand Kyber’s engineering and customer-facing teams and to productize its SDK for enterprise deployment. Lightspeed framed the investment as a bet on the infrastructure layer required for physical AI, arguing that software reliability and latency are fundamental to safe and scalable remote operations.
SDK Designed to Minimize Milliseconds of Lag
At the heart of Kyber is an SDK that prioritizes end-to-end latency and synchronization across media and controls. The software builds on video-streaming techniques to ensure operators see what devices see in near real time, while also adapting to each device’s compute limits and network conditions. Kempf, who previously worked on cloud gaming and is well known as the lead developer behind a widely used media player, emphasizes that even tiny delays can be consequential when commanding machinery in the physical world.
Open Source Core with Enterprise Productization
Kyber’s technical foundation is open source, a choice intended to broaden adoption and foster community-driven improvements. The company offers a commercialized, supported product for enterprises that require larger-scale deployments, complemented by consulting and hands-on support. Kyber also provides forward-deployed engineers who work directly with customers to tailor installations and ensure observability, mirroring service models used by other infrastructure vendors.
Target Verticals: Robotics, Drones and Remote IT
The startup is prioritizing three segments: robotics, drones of all types, and remote IT access. Kyber’s founders argue those use cases share a need for low-latency control where operators and compute resources are geographically separated from the machines they manage. In remote IT scenarios, demand has been especially strong because enterprises want secure, high-performance ways to access infrastructure without relying exclusively on legacy remote-desktop systems.
Commercial Deployments Across Defense, Telco and AI
Kyber reports active commercial deployments with customers in defense, telecommunications, robotics and artificial intelligence. These early projects range from remotely piloted systems to fleet management and secure remote access for critical infrastructure. Company representatives say the software is being stress-tested in demanding operational environments where observability and the ability to update distributed devices without physical access are essential.
Global Team and Forward-Deployed Engineers
The startup employs roughly 25 full-time staff and maintains offices in Paris, San Francisco and Singapore to support an international customer base. A significant portion of Kyber’s personnel are forward-deployed engineers who help with on-site integration, tuning and long-term maintenance. That hybrid model—open-source core plus in-person engineering services—aims to lower the barrier for organizations that lack bespoke tooling but require enterprise-grade reliability.
Scaling from Thousands to Millions of Devices
Founders highlight that while some companies have built in-house systems for remote control and teleoperation, those bespoke solutions are not designed to scale to very large fleets. Kyber positions itself as a scalable alternative that can handle everything from a handful of devices to potentially millions, an important distinction if autonomous agents and robotic fleets proliferate across urban and industrial environments. The company says observability and automated monitoring will become increasingly critical as scale grows.
Kyber’s funding round and product roadmap reflect a broader industry pivot toward “physical AI” infrastructure that reduces latency and improves reliability for real-world systems. By combining an open-source project with enterprise support and hands-on engineering, Kyber aims to accelerate deployment cycles for customers in multiple sectors while retaining flexibility for smaller users. As robotics and drone usage expands, the demand for low-latency, synchronized control and scalable device management is likely to grow, putting Kyber squarely in the middle of an emerging market.
