Between Two COO's with Michael Koenig

The episode delivers substantive technical insights about orbital data center architecture, particularly around packaging efficiency, radiator design, and why the second stage becomes the data center itself. However, significant portions consist of re-explanation of previously covered material and conversational filler that could have been tightened, reducing the novel insights per minute.

“By using the entire second stage of the rocket as one large radiator, we further improve the packaging efficiency because we're taking every kilogram of metal that we're taking into orbit. Really every kilogram of mass is itself being used in the service of providing this GPU compute.”

“the operating principle for the company is that if you look at what we're providing, which is AI compute, it only makes sense because the M, the end M compute user, needs to be indifferent as to whether that compute comes from outer space or whether it comes from a terrestrial data center”

The framing of orbital data centers as an economic problem rather than a pure physics problem is relatively fresh. However, the core thesis - that SpaceX-style reusability and cadence unlock economics - is increasingly familiar in startup discourse. The specific architectural choices (second stage as radiator) are more novel than the underlying philosophy.

“Our thesis a little bit different and a lot of it derived from the fact that Beijing was very much a, ah, business driven entrepreneur as AM M I. And so we're focusing on that and we're confident that our architecture gets us to an hourly GP cost that is competitive with the cheapest form of AI compute on the ground.”

“which is why our architecture contemplates that the second stage of the rocket is the data center. So there is no, there's no dispensing of smaller satellites.”

Joe Yaffe is genuinely credentialed: 31 years as a top Silicon Valley lawyer, COO and Chief Legal Officer of a venture-backed space company, and directly involved in building the business. He's a practitioner navigating real regulatory and operational problems at scale, not a consultant or commentator. His perspective is from inside the machine.

“31 years in large law firms or building a company like Robinhood. And what it actually takes to build a sustainable business.”

“I'm doing is both navigating that immense regulatory landscape and frankly working actively. And it's been very promising with senior folks at all of those agencies to come up with a more streamlined approach for commercial space at large.”

The episode lacks concrete numbers on most claims. While it mentions $365M raised, a December 2028 launch target, six-year GPU lifespan assumptions, seven-year Northern Virginia data center wait times, and megawatt-scale power, there are almost no specific numbers on costs, performance metrics, chip counts, or financial projections. The $2 billion comparison to infrastructure spend is vague. Many architectural claims are explained in principle but lack quantified validation.

“$365 million. Baiju, uh, seeded it. Real investors came in.”

“we have a timeline to launch our, what we call Mega, our 1 megawatt compute data center in December of 2028.”

The host asks sharp, probing questions that advance the conversation: the chip lifespan problem, redundancy failures in orbit, regulatory complexity, and the practical operating cadence. Follow-ups are substantive. However, the host occasionally accepts answers without pushing back or asking for more specificity (e.g., on economic modeling assumptions, on how redundancy actually works technically). Some softballs near the end on the 10-year vision.

“How do you plan for that and how do you think because SpaceX is carrying payloads up, whereas your payload is built into the rocket. Um, how do you think about those learning moments where, you know, things may not go according to plan?”

“if a part of one of the satellites dies in the middle of a customer's job, what happens? Whoa. Does the work jump to another satellite? Does it fall back to earth?”