CesiumAstro acquires Vidrovr to embed AI in communications systems

SAN FRANCISCO – CesiumAstro announced the acquisition Feb. 26 of Vidrovr, a startup that specializes in artificial intelligence for multimodal signals analysis.

Terms of the transaction, which closed in late 2025, were not disclosed.

CesiumAstro acquired Vidrovr to accelerate its campaign to embed AI in space telecommunications and intelligence, surveillance and reconnaissance infrastructure, enabling radio-frequency optimization, autonomous payload and satellite operations, and reconfigurable edge computing across product lines, according to the news release.

“Our systems must operate in an increasingly congested and contested environment,” Trey Pappas, CesiumAstro chief revenue officer, said in a statement. “By embedding AI directly into our telecommunications payloads, we enable adaptive RF optimization, autonomous tasking, and real-time decision-making at the edge. This reduces latency, improves spectrum efficiency, and allows our customers to operate resilient, self-optimizing space networks at scale.”

CesiumAstro, founded in 2017, builds software-defined phased-array communications systems for commercial and military space and airborne platforms. In 2024, CesiumAstro unveiled Element, a multi-beam active phased-array satellite.

CesiumAstro is growing rapidly, thanks in part to a $470 million funding round announced Feb. 2.

“We’re scaling the company as an industrial producer of this advanced technology,” Shey Sabripour, CesiumAstro founder and CEO, told SpaceNews in a recent interview.

The Vidrovr acquisition “positions CesiumAstro to further elevate its best-in-class digital processors and active phased-array technologies,” the news release said.

Onboard or Cloud Processing

Vidrovr technology will provide AI-enabled workload orchestration to allow “satellites to determine which data should be processed in orbit and which should be routed to ground-based cloud and enterprise systems,” according to the CesiumAstro news release.

Vidrovr was founded in 2016 by Joe Ellis and Daniel Morozoff-Abezgauz, then PhD candidates in Columbia University’s Digital Media and Multimedia Lab. Following the acquisition, Ellis, now Vidrovr chief technology officer, is overseeing the integration of machine-learning capabilities across CesiumAstro’s product portfolio, with a focus on next-generation, AI-native space systems.

“By embedding analytics and autonomy directly into the Element family of satellites and communications payloads, CesiumAstro is establishing a real-time planetary intelligence layer” to “route the necessary data intelligently across an expanding network of space-based assets,” Ellis said in a statement. “Our goal is to bring machine learning inference as close to the data as possible, on orbit, and route the most important data as quickly as possible to where it should be processed on Earth.”