Press & media
Everything you need to write about Storykept: boilerplate, a fact sheet, founder background, story angles, an FAQ, downloadable images, and a media contact. Interviews and demo access available on request.
In one paragraph
Storykept is a voice-first, AI-guided app for keeping a family’s stories. People record memories by voice — alone, by inviting a relative through a private link, or on a browser-based recorded call — and an AI listens, asks gentle follow-up questions, and weaves contributions from several family members into one living, searchable archive that keeps every original recording. Stories can be composed into a designed digital book and downloaded as a PDF. Storykept is private by default, stores data in European data centres, never trains AI on users’ content, and is built to last for decades. The interface is available in 34 languages. Storykept is based in Austria.
You may reproduce this paragraph verbatim. Short version: “Storykept is a voice-first AI that turns a family’s spoken memories into one living archive — kept in their own voices, and exportable as a downloadable digital book.”
Announcements
Available now · for immediate use
Storykept’s beta is open at storykept.app. Anyone can record a first memory by voice and watch the AI turn it into a written story that keeps the original recording. Journalists and reviewers can request free access and a walkthrough at the contact below.
Embargoed · for release [DATE], autumn 2026
[CITY], Austria — [DATE], 2026
Storykept, a voice-first app for keeping family stories, today opened to the public. Instead of asking people to write a memoir, Storykept lets them simply talk: an AI interviewer asks gentle follow-up questions, turns the recording into faithful prose, and connects the people, places and dates inside it — building a private, searchable archive a family can add to for years and compose into a downloadable digital book.
Its defining feature is multi-voice memory. When several relatives contribute to the same story, Storykept weaves their accounts into one narrative that keeps each person’s version — and every original recording — rather than flattening them into a single telling. Storykept keeps recordings untouched and never clones or synthesises anyone’s voice.
Storykept is private by default, stores data in European data centres, and does not train AI on users’ content. The interface is available in 34 languages, with story capture and writing supported in several, so multilingual and diaspora families can keep memories in the language they were lived in.
“Most people will never sit down and write a memoir — but they’ll happily talk for an hour. I wanted the software to do the hard part, so the only thing a parent or grandparent has to do is tell the story.” — Georgi Kostov, founder
Storykept is available at storykept.app with a free trial and one €59 annual Core subscription, including the private archive and digital book/PDF export.
Bracketed items are placeholders to confirm before release.
Fact sheet
Founder
Storykept was founded and built by Georgi Kostov, a solo technical founder based in Austria. He built it after watching how easily a generation’s stories disappear — parents and grandparents with a lifetime of memories who would never sit down to write them, but would happily talk. One of Storykept’s earliest users records in Bulgarian, which is why the product treats languages beyond English as first-class from the start.
Quotes you may use:
“We keep every original recording, untouched. The point isn’t to replace a voice with AI — it’s to make sure that voice is still here in twenty years.”
“A book is finished once. A family’s story is never finished — so we built an archive that grows every time someone adds a memory.”
For journalists
A few angles to pitch your editor — pick the one that fits your readers.
Stories that outlast us
AI used to fight digital decay — capturing the memories a generation is about to take with them, and keeping them readable and hearable for decades.
The anti-deepfake stance
A timely AI-ethics angle: Storykept keeps real recordings and refuses to clone or synthesise anyone’s voice — the opposite of voice-cloning tools.
Memory across languages
Built for multilingual and diaspora families: an interface in 34 languages and capture in several, so a story is kept in the language it was lived in.
Many voices, one digital book
Friends and relatives contribute their own versions, and Storykept weaves them into a designed digital book the family can download and share.
Aging, care, and connection
Capturing elders’ stories and connecting generations — a family and caregiving angle around reminiscence and presence.
Privacy-first, European AI
A consumer AI product built EU-first: data in European data centres, no training on user content, private by default.
Anticipated questions
Isn’t this just ChatGPT?
No. Storykept is a purpose-built pipeline for spoken memory: it transcribes speech, runs a guided interview with follow-up questions, composes faithful first-person prose, tags people and places, and merges multiple contributors with attribution. The conversation is designed for storytelling, not a general chatbot.
Do you train AI on people’s stories?
No. User content is never used to train AI models. Content is processed only to fulfil the user’s request, under data-processing agreements.
Do you clone or synthesise voices?
No — and this is a permanent commitment. Original recordings are kept unaltered; Storykept never creates a synthetic version of anyone’s voice.
What languages does it support?
The interface is available in 34 languages. Story capture, transcription and writing are supported in several languages — English and Bulgarian are fully supported at launch, with more rolling out after native-speaker review.
Who is it for?
Adult children capturing their parents’ and grandparents’ stories; anyone keeping a family’s memory; and multilingual and diaspora families keeping stories across languages and borders.
Where is data stored, and can people get it out?
Data is stored in European data centres and is private by default. Users can export everything — text, structured data, and original audio — at any time.
What does it cost?
The free trial includes about one Story Hour. Core costs €59 per year and includes 24 pooled Story Hours, the private archive, and digital book/PDF export.
When and where can I try it?
The beta is open now at storykept.app; the public launch is planned for autumn 2026. Reviewers can request free access and a walkthrough.
Downloadable
Free to use in coverage of Storykept, with credit “Courtesy of Storykept.” Lifestyle images are illustrative. Need a specific format, a higher resolution, the founder headshot, or product screenshots? Email the contact below.
Get in touch
Georgi Kostov — founder, Storykept
Email: hello@storykept.app
Web: storykept.app
Interviews, product demos, and free review access available on request. We typically reply within one business day.
About Storykept: a voice-first, AI-guided app for keeping a family’s stories — recorded by voice, woven into one living archive kept in their own voices, and exportable as books. Private by default, European-hosted, built to last. Based in Austria.
