Baiju Bhatt left Robinhood to build a space-based solar power grid — first satellites launch this summer

Summary

Baiju Bhatt, co-founder of Robinhood, left the company to build AetherFlux, a startup whose core mission is constructing a power grid in orbit. Nearly two years in, the company has roughly 30 employees, a functioning clean room, and its first satellite in active assembly. Two demonstration satellites are scheduled to launch this summer, with a public event planned — described as somewhere between a scientific conference and a desert gathering — to showcase live power beaming from space.

The Core Technology

AetherFlux places solar panels in a dawn-dusk, or 'terminator,' orbit, where satellites track the boundary between sunlight and shadow and remain illuminated by the sun nearly continuously. Ground-based solar panels produce peak output for roughly one to two hours daily. The same panels in this orbit generate power around the clock at near-peak levels, fundamentally changing the economics of solar energy collection.

The company is pursuing two applications of this orbital power. The first beams energy down to locations without grid infrastructure, with Department of Defense forward-deployment scenarios as the primary near-term customer. The second, and strategically more significant, involves hosting compute directly in orbit, eliminating transmission losses by placing chips where the power already is.

The Space Data Center Thesis

Bhatt credits Jeff Bezos with catalyzing mainstream interest in space-based data centers after Bezos suggested last fall that the concept could materialize in the 2030s — prompting a wave of investors and operators to argue it could happen within five years. AetherFlux is positioning itself as the power infrastructure layer for that compute.

The company's near-term compute roadmap targets a satellite with 10 to 20 kilowatts of electrical power supporting a cluster of 8 to 10 interconnected GPUs capable of hosting a model and running inference workloads. Bhatt frames inference as the realistic starting point, noting that the orbital mechanics of connecting 100,000 GPUs at scale remain a meaningful unsolved problem. The broader vision scales toward constellations delivering gigawatts of orbital power within a decade.

The original space-based solar power concept, which called for microwave-beaming satellites one to five kilometers in diameter, was a deliberate dead end. AetherFlux's approach consciously sizes its demonstration missions to what is commercially fundable and technically de-risked.

Team and Funding

The company has secured enough funding to operate a full lab facility, maintain liquid nitrogen infrastructure, and run parallel satellite assembly. The team includes engineers from JPL, SpaceX, and Lawrence Livermore National Laboratory, alongside optics specialists. Bhatt used a combination of physics-grounded spreadsheet modeling and AI-assisted technical self-education to validate the underlying economics, explicitly asking at each component whether raw manufacturing cost versus purchase price justified vertical integration.

Execution Philosophy

Bhatt draws a direct analogy between his Robinhood thesis — that financial services would become the next native vertical on mobile — and AetherFlux's thesis that energy is the next native vertical on low Earth orbit as a platform. He cites Varda Space as a tactical model, noting that Varda's willingness to get hardware into orbit quickly, accept imperfection, and iterate compressed the credibility timeline significantly. Will Bruey at Varda served as an early informal advisor, whom Bhatt has described as his 'space Sherpa.'

Bhatt is candid that building a hard-tech company after Robinhood has not been easier — single-function teams, personnel turnover, and the long gap between technical progress and commercial revenue all compound. He describes the venture as 'enterprise SaaS on absolute hard mode.'