About

Practice should be private.

A few times a year, every adult has a conversation that decides something — a job, a relationship, a home, a custody arrangement, a goodbye. People prepare for those conversations alone, in their head, or not at all.

Razen exists to give people a room to practise out loud, with a patient partner, until they're ready. That's the whole product.

What we're not

We're not therapy. We're not a coach. We're not a substitute for a human friend or a professional. We are deliberately narrow — a voice, a scenario, a debrief, and a way to run it again.

How we build

We use licensed neural voices that run on your device, and Voice Studio lets you design a fully synthetic voice of your own — Razen never records, uploads, or clones your real voice, and generated audio is watermarked. We ship the smallest possible set of features and we say no to a lot. We list every sub-processor publicly. We're registered with the UK ICO. We comply with the EU AI Act's transparency obligations from day one.

Where we are

Built in London. Available in the UK, the EU and the US at launch. Audio for EU and UK users is processed in EU regions; US user data stays in US regions. Sub-processors and data flows are listed at /legal/sub-processors.

Working with us

We're a small team. If you're a writer, designer, voice actor, lawyer, or coach with something to say about how this product should evolve, please write — razenai@outlook.com.