Y CombinatorBringing Supersonic Flight To The People
Y Combinator
Rolls-Royce Breaks Away 0:00
The team first tried to outsource a semi-custom supersonic engine to Rolls-Royce. Blake Schol says that was one of his worst decisions. It seemed practical at the time because an engine designer on the small team thought Rolls’s engines were easiest to adapt, and the industry expected a big-name supplier. The deal reached a Rolls board meeting in early 2022, then collapsed when the CEO was fired and no one wanted to approve anything.
Building The Engine In House 7:30
After American Airlines and then United placed pre-orders, Rolls broke with Boom publicly, and the company became hard to fund or partner with. Schol then chose to build the engine itself, with a former Pratt and Whitney supersonic team in Florida, a small GE role, and a maintenance shop. That decision later made boomless cruise possible. A Saturday simulation showed the production plane’s own engine had solved sonic boom months earlier.
Power turbine pivot 14:01
A March 2025 shock changed the plan. News of an AI power shortage and long turbine backlogs made the team see a first revenue product in the engine they were already building. A supersonic engine without its fan becomes a 42 megawatt turbine, close to the right size for data centers. That meant modular units, the same logic as blade servers, and a product they had already spent years developing.
Funding and old industry 18:30
Before that shift, the company was living paycheck to paycheck. Investors feared hard tech, and aircraft startups did not fit venture capital patterns. The speaker says product-market fit in hardware cannot be proved by shipping the way software does, which made funding harder. Older turbine and airplane makers also looked stuck. GE’s midsize turbines still come from a 1971 jet engine, and Boeing has not designed a new airplane since 2004.
Cabin tradeoffs 28:31
Design and engineering were worked together on the interior, but the first shape cost 1,000 miles of range on a 4,000-mile airplane. That forced a wider search for savings elsewhere. The answer came from designing the engine and airframe together, rather than fixing the engine first and fitting the plane around it.
Range and comfort 31:00
A dozen buildable engines were modeled with software called Makeboom, which wrapped an airplane around each engine and ran full flight simulations. One clear winner recovered most of the lost range. That joint optimization changed the cabin too. It helped deliver no middle seats, five-minute boarding and deplaning, Starlink, and an antenna built in so it would not add drag.
Defense and manufacturing 42:00
Missile shortages lead into a wider point about defense production. The answer is a flexible commercial base, like World War II factories that could switch fast. Today aerospace parts can take 18 months in the US. The goal is to move from a design idea to a finished turbine part in 24 hours and pivot quickly if needed.
AI-generated summary. It can be wrong or incomplete - check anything that matters against the original.
