Build the interface that listens.
We're not listing open roles yet — but we're always glad to hear from people who care about voice, AI, and craft.
Open roles
None right now. When that changes, they will be listed on this page rather than only on a job board — so this is the page worth checking, and the one that will always be current.
What we're building
WisprBee is a voice keyboard and a meeting notetaker for the Mac. Holding a key and speaking has to produce text good enough to send, in whatever app is in front of you, in under two and a half seconds. That constraint touches almost everything: audio capture, macOS accessibility APIs, model selection and latency, and a lot of careful judgment about what the software should and should not change about somebody's words.
How we work
Small team, short chain between deciding and shipping, and a strong bias toward being honest in public about what the product does not do yet — the about page is written that way on purpose. We care about craft in the parts users never see, and about not shipping claims we cannot back.
Get in touch anyway
If you work on speech, on-device inference, macOS internals, or interface design and this sounds like your kind of problem, write to us. Tell us what you have built and what you would want to work on here — a link to something real is worth more than a CV.