
The Idea: Akshay Narisetti had already built one AI note-taker before this one. As a founding member of Omi, one of the world's largest open-source note-taking devices, he had a front-row seat as the category filled up. Otter, Fireflies, Granola, Zoom, Read AI, every meeting tool on earth bolting on a bot that joins your call and emails you a summary afterwards. They were all solving the same meeting: the one that happens on a screen. Narisetti's observation was that most conversations don't happen on a screen. The sales call in a car park. The consultation. The site walk. The open house. The part of the client dinner where they finally tell you what they actually want. AI needs enormous context to be useful, and most of that context lives offline, where no bot can dial in. The existing answer was hardware, and the existing hardware asked a lot: wear a pendant, clip on a pin, pay a subscription forever. Narisetti and co-founder Gabriel Dymowski went duller and cheaper instead. A credit-card-shaped puck that magnets onto the back of the phone you're already holding. This was the beginning of Pocket.
The Execution:
The lesson? Narisetti didn't find a gap in the market, he found the market everyone had walked straight past. Half a dozen well-funded companies were fighting over the video call. Nobody had built properly for the conversation happening in a car park. Pocket's edge was never better AI, it was picking a customer the software players structurally couldn't reach, then charging that customer $129 once instead of a subscription they'd have to justify every month.