Endeavor Optical Networks raises $10.75M to route data-center traffic through satellites

Why it matters

EON is testing whether a 20-satellite network can give hyperscalers dedicated intercontinental capacity faster than new subsea cables can be built.

Endeavor Optical Networks emerged from stealth on August 4th with $10.75 million in seed funding and a plan from co-founders Charlie Horowitz and Tyler Presser to carry intercontinental data traffic through laser-equipped satellites.

The two founders are pairing experience from opposite sides of the problem. Horowitz, EON's CEO, began his career at Palantir and later became the 13th employee at satellite manufacturer Apex. Presser, EON's CTO, is an astronautical engineer who worked on deep-space mission design before starting EON in May.

Their proposed service is built for hyperscalers, AI labs and other large infrastructure operators, rather than households buying satellite internet. EON wants customers to lease or own dedicated routes between data centers, with a satellite in medium Earth orbit relaying optical signals between ground stations as much as 10,000 kilometers apart.

TechCrunch reported that General Catalyst and Andreessen Horowitz backed the seed round. General Catalyst partner Jeannette zu Furstenberg led the investment, while Apex CEO Ian Cinnamon invested personally. EON did not disclose its valuation.

The capital buys EON time to prove that the founders' proposed data rate can survive the transition from a presentation to orbit. EON plans to build an optics lab, add engineers and run ground tests before launching a demonstration spacecraft around the end of 2027. Horowitz told TechCrunch that the demonstrator is intended to deliver between 800 gigabits and 1 terabit per second. EON's initial commercial target is 2.4 terabits per second.

A software operator and a mission designer

Horowitz arrived at Apex from a software background and worked first as chief of staff before becoming director of special projects. In a 2023 interview with Apex, he described hardware schedules as fundamentally harder to manage because delays extend through suppliers and physical components, rather than remaining inside an engineering organization.

That experience matters to EON's approach. Horowitz told TechCrunch that EON intends to concentrate its engineering on the optical communications terminal and buy an existing satellite bus, potentially from Apex. TechCrunch reported that EON would direct spending toward specialized components, including gimbals used to point the laser.

Presser brings the mission-design side. He earned joint bachelor's and master's degrees in astronautical engineering from the University of Southern California in 2021 and later pursued a PhD there. His USC research profile describes work on rapid spacecraft trajectory design, automated maneuvering and low-thrust transfers in cislunar space. Before EON, he led mission formulation at Advanced Space.

EON has also recruited operators who know the two terrestrial systems it must connect. Head of Lasercom Wes Baxter previously worked on optical communications at NASA's Jet Propulsion Laboratory and Amazon's satellite program, according to EON's team page. Head of Growth Michael David Francois spent approximately 15 years at Google working on infrastructure that included subsea networks.

Selling routes instead of general coverage

EON's small proposed fleet reflects the commercial model. Horowitz and Presser envision about 20 satellites, each supporting dedicated links between continents, with the initial group providing continuous coverage for early customers. Ground stations would be placed in multiple locations so EON could steer traffic around clouds and unfavorable weather.

The founders are focusing on routes where cable capacity is scarce, expensive or slow to add. TechCrunch cited France to Australia and Africa to South America as examples. EON says it is speaking with hyperscalers and AI labs, though it has named no customers or signed contracts.

EON is pitching deployment speed and control. Subsea cables require ships, surveys, landing stations and permits across several jurisdictions. EON says an optical route could be activated in months and claims it could deliver capacity at one-tenth of the cost per bit of conventional long-haul options. Those figures remain projections. EON has yet to fly hardware, publish an operating cost model or demonstrate a commercial link.

Dedicated capacity could still give EON a narrower entry point than a general broadband network. A hyperscaler may value a route that avoids shared conduits and cable landing stations as a backup path, even if orbital links do not replace the vastly larger capacity already moving through subsea fiber. EON also plans to offer routes for bulk replication between data centers and for connecting new compute regions before terrestrial infrastructure is ready.

The atmosphere sets the test

Laser communications can carry data at far higher rates than many radio systems, but the beam must be pointed precisely and then pass through an atmosphere that can distort or block it.

NASA's communications guidance identifies pointing accuracy, spacecraft power and thermal constraints, atmospheric distortion and cloud attenuation among the core problems. Heavy cloud cover can stop an optical connection entirely, forcing a network to wait or redirect traffic to another ground station. Higher data rates can also require adaptive optics to correct the incoming wavefront.

EON's redundancy plan addresses weather by distributing ground stations and using forecasting to select an available path. It also raises the amount of ground infrastructure and coordination required to provide the quality guarantees that data-center operators expect.

The sector's record calls for a measured view of EON's targets. A Government Accountability Office review found that the Space Development Agency's laser communications program had suffered delays and had moved into later satellite tranches without completing its planned space-based mesh demonstration. The report covered a different architecture and government procurement program, but it documents how difficult acquisition, pointing and interoperability remain even with substantial budgets.

A race with hardware already in orbit

Kepler Communications launched its first tranche of 10 optical relay satellites in January 2026, beginning commissioning of a network aimed at moving data among spacecraft and ground systems.

Blue Origin's TeraWave is targeting enterprise, data-center and government customers with proposed speeds of as much as 6 terabits per second. TechCrunch described its planned constellation as larger than EON's roughly 20-satellite network.

EON's opening comes from staying focused on a specific transaction: one customer buying a high-capacity path between two points. A smaller constellation could require less capital and reach service sooner, provided EON can establish reliable links through weather, build enough ground-station redundancy and meet the service levels expected of terrestrial fiber.

Horowitz told TechCrunch that EON faces "no physics problems." The phrase describes a decision to assemble known components around an aggressive engineering target, rather than depend on an undiscovered scientific breakthrough.

The physics is understood. EON's work begins with making it reliable, affordable and repeatable at a throughput that data-center operators will buy.