What Irth collects, and why.

Who is responsible

Irth Studio runs this site and the archive behind it, and is responsible for the information described here.

The quickest way to reach us about anything on this page is the contact form on this site. It reaches the team directly and we answer from there.

What we collect

Only what a person types, records or uploads. We buy no data from anyone, and nothing is gathered in the background while you read the site.

• Application details — your full name, email address, city, phone number if you give one, the language you prefer and the sector you work in.
• Your intro recording — the sixty-second video or audio you record or upload when you apply. It carries your face and your voice, and it is the source we build your voice clone from.
• Consent evidence — when you tick the consent box we store the exact wording you agreed to, the moment you agreed, and the IP address the agreement came from. It is kept as proof of consent, not to follow you around.
• Your story material — recordings, transcripts, summaries, the answers you write, and any reference photographs you upload together with the descriptions you write for them.
• Messages — anything you send through the contact or sponsorship form: name, email, phone, organisation and the message itself.
• Waitlist entries — email, name, role, city and what brings you to Irth.
• A one-way hash of your IP address, held briefly so the public forms cannot be flooded. The address itself is not stored for this.

Why we hold it

We do not sell your information and we do not use it for advertising. Your material goes to the production providers listed below only to make your own story.

• To review your application and tell you the outcome.
• To produce your documentary, your written record and the shorter pieces that come with it.
• To contact you about your own story, your account, or a message you sent us.
• To prove that consent for voice cloning and story use was given, and what it said.
• To keep the public forms working — rate limiting, abuse prevention, basic security.

Who else handles it

Turning a recording into a film means the material passes through specialist providers. Each one receives only what its own step needs.

• Supabase — the database, the file storage and the sign-in system behind the whole platform.
• Vercel and Railway — hosting for the website and for the production service.
• Google Cloud Speech-to-Text — receives your recording to turn it into a transcript. We send it to Google's European region.
• OpenRouter — receives your transcript and story text and routes it to the language models we use (today, models from Anthropic, OpenAI and Google) to draft the treatment, the scenes, the article and the captions.
• ElevenLabs — receives the audio of your voice to build the clone you consented to, and the narration text it then speaks in that voice.
• fal.ai — receives generated images and prompts to produce the video shots and the music bed of your film.
• Google Cloud Text-to-Speech — the fallback narrator when a cloned voice is not used. It receives the narration text.
• Resend — delivers our email to you, so it receives your address and the contents of that message.
• Several of these providers process data outside the Kingdom, including Supabase, Vercel, Railway, Resend, OpenRouter, ElevenLabs, fal.ai and Google. Each is engaged under its own data-processing terms.

What stays private, and what is public

Cookies and what the site keeps in your own browser are covered on the cookies page.

• Private — your intro recording, your raw audio, your reference photographs, the internal script and the treated voice master. They sit in private storage that only the Irth team can open, through short-lived signed links.
• Public — the finished film, and the images and quote cards generated with it, are served from public storage: anyone with the link can open them. That is the point of an archive, but it deserves saying plainly.
• Nothing is published without a human review and the leader's own approval.

How long we keep it

The archive is built to last, so a published record is kept indefinitely unless you ask us to take it down. Applications that are not approved, waitlist entries and form messages are kept only while they remain useful to the review, and are removed once they are not. You can ask us to delete yours at any time through the contact form.

The hashed IP addresses used for rate limiting fall out of use within the hour.

Your choices

Send any of these through the contact form on this site. Whatever rights the law gives you locally sit on top of this list, not instead of it.

• Ask for a copy of what we hold about you.
• Ask us to correct anything that is wrong.
• Ask us to delete your application, your account, or a recording that has not been published.
• Withdraw your voice consent — we delete the cloned voice and stop narrating in it. Work already published can be taken down at your request.
• Ask a question about any of this and get an answer from a person.

Children

Irth is built for adults telling the story of their own working life. We do not ask for information from children and we do not knowingly keep any. If a child has sent us something, tell us and we will remove it.

Changes to this notice

If this notice changes substantially, the date at the top changes with it, and we tell anyone already in the archive by email before the change takes effect.