Pocket Raises $11M as Open Vision Engineering Expands AI Hardware

Pocket, the flagship product from San Francisco-based Open Vision Engineering, has raised $11M in new funding from Accel, Y Combinator, and a group of technology operators that includes Vercel Founder and CEO Guillermo Rauch, ElevenLabs Co-Founder Mati Staniszewski, and Opendoor CEO Kaz Nejatian. Led publicly by CEO Akshay Narisetti, the company is building an AI-native conversational capture hardware device designed to record conversations, generate transcripts and summaries, and turn spoken work into usable follow-up actions. The backer list matters because it combines venture capital with operators who understand how new interfaces become daily work habits.

The funding will support hiring across design and engineering, expansion of the Pocket software platform, and exploration of additional hardware form factors. The announcement lands as AI hardware tries to carve out an identity separate from smartphones and laptops, with products designed around specific workflows instead of another general-purpose screen. More importantly, this round points to a growing belief that AI advantage is moving from model access alone toward products that remove friction from everyday work.

What Happened

Open Vision Engineering secured $11M to accelerate development of Pocket, a dedicated conversational capture device that records discussions and turns them into transcripts, summaries, action items, and follow-up material through companion software across mobile, desktop, and web. The announcement did not identify a named round type in the primary source materials reviewed, but it did disclose the amount, the funding date, the CEO quote, the named investors, and the company's intended use of proceeds.

The investor syndicate combines institutional backing from Accel and Y Combinator with operator capital from Guillermo Rauch, Mati Staniszewski, and Kaz Nejatian. That mix gives the round a different signal than a generic AI hardware headline. It suggests investors are not just backing another device, but a workflow thesis around how conversations become structured knowledge.

Pocket approaches AI hardware from an intentionally narrow angle. Instead of trying to become another all-purpose consumer device, it focuses on a recurring problem: preserving the conversations that usually vanish after meetings, calls, brainstorms, interviews, and customer discussions. That focus may sound simple, but simplicity has become surprisingly rare in a market addicted to bigger surfaces and louder notifications.

Why This Matters

The AI market has entered a phase where model performance is no longer the only story. Foundation models are increasingly accessible, which makes product execution, interface design, distribution, and workflow ownership more important. Pocket reflects that shift by treating the device as an interface for capturing real-world context, not as a novelty gadget looking for a use case.

The product is built around the idea that spoken work is still under-instrumented. Teams spend hours in conversations where decisions, objections, context, and next steps appear briefly and then disappear into memory. Pocket is trying to make that information usable without asking people to stare at another tab, invite another meeting bot, or turn every discussion into an administrative chore.

That distinction matters because information has never been the scarce resource. Attention is the scarce resource, and the best productivity products often win by making useful behavior easier rather than asking users to become more disciplined. If Pocket can keep the capture experience simple while making the output valuable, it has a cleaner wedge than broad AI tools that promise to do everything.

Market Context

Investor interest in dedicated AI hardware has grown alongside demand for workflow-specific AI products. The first wave of generative AI centered on chat interfaces, copilots, and software layers that could be added almost anywhere. The next phase is starting to ask where AI should live when the job is not typing into a box, but capturing context from the physical world.

Pocket fits directly into that question. According to the company's funding announcement, more than 10,000 devices shipped on launch day in October 2025, more than 35,000 devices had shipped by March 2026, and the business reached a $27M annualized revenue run rate while sustaining several months of more than 50% month-over-month revenue growth. Those figures give investors something stronger than a category narrative: evidence that users are already adopting the workflow.

Venture firms rarely fund momentum alone. They fund evidence that customer behavior is changing, and Pocket's early numbers suggest users may want AI hardware when it solves a narrow, repeated problem. The market lesson is not that every AI product needs a device; it is that some AI experiences may work better when the interface is purpose-built for the moment it serves.

Competitive Landscape

Open Vision Engineering is entering an active segment where AI software companies, meeting tools, voice recorders, and hardware makers all want to own conversational intelligence. Many existing products rely on smartphones, meeting bots, or software integrations to capture conversations. Pocket's position is different because the hardware is dedicated to ambient conversational capture rather than bundled into a general-purpose device.

That matters more than it first appears. Dedicated devices can reduce the friction of starting a capture session and remove the distractions that come with phone-based workflows. Instead of asking users to remember another app, browser extension, or meeting assistant, Pocket is trying to make capture feel like a physical habit.

The company also appears to be emphasizing interoperability rather than a single-model identity. A model-agnostic approach gives the product flexibility as foundation models continue changing, while keeping the customer value centered on outcomes: better notes, clearer summaries, action items, follow-up drafts, and searchable knowledge from conversations.

What This Signals

The Open Vision Engineering funding round illustrates a broader trend across venture capital. Investors are becoming more selective about AI companies that simply wrap existing models with incremental features, and capital is flowing toward teams that pair intelligence with differentiated product design and observable customer behavior. That is the difference between a demo and a company.

Durable technology companies rarely emerge from feature lists alone. They emerge when founders identify an overlooked friction point and remove it so effectively that users start wondering why the old process lasted so long. Pocket's wedge is not that conversations are important; everyone already knows that. Its wedge is that most conversations still produce too much lost context for how modern work actually operates.

The funding gives Open Vision Engineering more room to expand engineering and design capacity while exploring additional hardware categories. Whether those future products work will depend on execution, not narrative. Venture capital can accelerate a product that already has pull, but it cannot manufacture the daily behavior that makes a device worth carrying.

The Bigger Industry Shift

AI is entering an era where interface design may become just as important as model intelligence. For years, the market conversation revolved around who built the biggest model or shipped the broadest assistant. Increasingly, the better question is who can make AI disappear into everyday life without creating more complexity.

That shift benefits companies building focused experiences instead of universal platforms. Open Vision Engineering is betting that conversations are one of the richest sources of untapped organizational knowledge, and that capturing them naturally can create value long after the meeting ends. The $11M financing suggests investors see similar potential.

The next chapter for AI may not belong only to companies that ask users to spend more time staring at screens. It may also belong to companies that let people forget the technology is there while making the work around them easier to remember, search, and act on.