Why we exist
Most dictation tools are built for consumers — they ship audio to a third-party server, train models on what you said, and ask you to trust a privacy policy. That's a non-starter for a law firm, a hospital, or a fund.
SpeakrStream runs the entire stack on the user's own device. Audio is never uploaded. Transcripts are never stored centrally. The IT team can verify all of this with a packet sniffer in under five minutes.
Who we are
Founded in 2024 in London by ex-DeepMind and ex-Palantir engineers. Backed by a small group of operator-investors from the legal and financial services world.
Where we are
Headquartered in London. Data residency in the UK and the EU. SOC 2 Type II and ISO 27001 audited annually.