A profile of the founder who quit, came back, and shipped

The kitchen, in October 2024, has white tile to the ceiling and a brass tap with a leak the landlord has been promising to fix since June. There are two espresso cups on the table, both empty. Outside the window the leaves on the plane tree are doing what plane trees do in October. On the table, between the two espresso cups, is a printed sheet of paper. The founder who is leaving has put it there. The founder who is staying is reading it.

I am sitting in the same kitchen eighteen months later, in a different October, with the same founder who left and the same co-founder who stayed. The leak in the tap has not been fixed. The company they founded together has just shipped the product the printed sheet of paper, in October 2024, was about. They are trying to remember which of them put the sugar bowl out that morning. Neither of them can.

The disagreement

The disagreement, as she describes it now, was not about the product. It was about the timeline for the product. "I thought we needed two years to build the right thing," she says. "He thought we needed nine months to ship the wrong thing and then iterate." She pauses. "We were both right and we were both wrong."

The co-founder, when I ask him the same question separately three days later, says something different. "She thought we should slow down. I thought slowing down was the failure mode. The thing we couldn't agree on was what we owed the team." He pauses for longer than she did. "I think she was right and I think she was wrong, but I don't think the answer is in those words."

"We were both right and we were both wrong."

The eighteen months

She left, in October 2024, and went to Berlin. She worked for six months at a research lab she had been adjacent to for years. She wrote, she says, "the kind of code I had not written in three years." She published two technical papers. She did not, by her own account, plan to come back.

He stayed, in London, and he raised the company's Series B in March 2025. The round was led by a fund she had been on the board of for an investment that pre-dated the company. The fund's GP called her, in March 2025, to tell her about the lead. "It was the most awkward phone call of the year," she says. "I told him to do it. I meant it." She is, when she says this, looking at the brass tap.

In July 2025 she came back. Not to the role she had left — that role no longer existed in the org chart he had built without her — but to a smaller role with a clearer mandate. The role was Chief Product Officer. She had been a co-CEO. The transition is the part of the story that, when I ask both of them about it on different days, neither of them describes the same way.

The product

The product, which shipped on April 11, 2026, is a developer tool I am not going to name in this profile because the editorial agreement she made with me when she agreed to this piece included a request not to make it about the product. The product is good. It has been adopted by the kind of teams that adopt this kind of product. The metrics are private and on the right side of zero. None of that is the story.

The story is that they shipped it. The story is that they shipped it together. The story is that they were both right, in October 2024, and they were both wrong, and somewhere between the leaking brass tap and the kitchen with the same sash window eighteen months later, they figured out that the thing they were arguing about was not what to build but how to build it together.

The chief of staff comes through the kitchen at 11:14, looking for an iPad cable. She finds one on the counter. She nods at us and goes back into the office. The co-founder watches her go. He turns back to me. He says, "We almost didn't."