The way we work is changing.
Our problem
As more of our work is delegated to AI, the precision and depth of our own input become ever more important.
Our work shifts toward deciding what we want, connecting loose ideas, and provide enough supporting context for AI to deliver the outcome we expect.
Keyboards have become the bottleneck. Despite AI is particularly good at making sense of that kind of input, we are naturally reluctant to type out long, unstructured blocks of text without first organizing or editing them.
To get the most out of AI systems, we need an input with higher throughput to feed snippets of ideas and context as quickly and naturally as we can think of them.
Voice gives us the bandwidth we need.
We could start speaking before we knew exactly how to phrase an idea, develop it as we think out loud, and provide far more context without turning every thought into a sequence of keystrokes.
Our context should move with us.
Just as digital wallets gave us one place for different payment and membership cards, we believe voice needs a shared layer across the devices and software we use.
One that understands what we are working on, where our words are headed, and how we communicate in each context.
Our goal is to go beyond dictation and make voice a practical, everyday interface for interacting with AI agents and other people.
Our team
We build products that we wish existed.
Several of us previously worked together at the London-based startup Nothing, where we shipped a dozen smartphone and audio products used by tens of millions of people every day.
We were obsessed with building experiences people would genuinely love to use.
Being tech enthusiasts means we rarely accept an experience just because that is how it has always worked.
We are constantly trying new products, comparing approaches and noticing where an experience feels unnecessarily tedious or limiting. Many of our strongest product convictions start there.