What MemoriesBox does

What MemoriesBox is

Most services in this category do one of two jobs. They send questions and turn the answers into a book, or they build a chatbot after somebody has died. MemoriesBox does a third thing, and the order matters.

A person sits down, on their own schedule, and is interviewed. Not with a list of fixed prompts, but with questions that follow from what they just said. That happens across many sessions over months. The audio is kept as audio. After about 30 minutes of recording there is enough material to build a model of that person's voice. Later, their family can ask the archive a question and hear an answer in that voice, in English, Russian, Ukrainian or Spanish. Everything sits in the European Union, and everything can be exported to your own drive at any time.

The point is not the technology. It is that 47% of Americans regret not recording or documenting a conversation with someone who has since died, and 59% have never done it at all (YouGov, June 2022, n=6,090). That survey covers recording or documenting, so it is not a statement about voice specifically. It is still the shape of the problem. Almost everyone means to. Almost nobody does.

The nine things it does

How an archive gets built

1. Someone starts. Usually an adult child, sometimes the storyteller themselves. The first 14 days are free.
2. The first session runs short. Twenty minutes is enough. The aim is to get past the self-consciousness, not to cover a decade.
3. Sessions accumulate. A month in, there are several hours of audio and a rough shape: childhood, work, the move, the marriage, the years nobody talks about.
4. The voice model becomes possible. About 30 minutes of recording is the threshold. It is usually crossed in the first two or three sessions, long before the story is finished.
5. The story gets organised. Sessions become chapters, photographs attach to the parts they belong to, and the gaps become visible.
6. The storyteller decides who gets a key, and to which parts, while they are alive and able to say so.
7. Later, someone asks a question. The archive answers from what was actually recorded, in the person's own voice, labelled as AI.

Step seven is the part that does not exist elsewhere. Step six is the part that makes it defensible.

What it will not do

• It will not produce a person. What is kept is a recording that can be asked questions. It is not a continuation of anybody, and we will not write a sentence suggesting otherwise.
• It will not invent. The archive answers from what was said. It does not generate opinions the person never expressed, and it does not fill silences with plausible material.
• It will not build anything from someone who has died. If there are no recordings the person made themselves, there is nothing for us to work from. We do not reconstruct people from photographs, messages, or a relative's description. See /trust/consent.
• It will not pretend to be a person during a conversation. Every conversation is labelled as AI, which EU AI Act Article 50 has required since 2 August 2026. See /trust/eu-ai-act-article-50.
• It will not chase you. No streaks, no push notifications, no nudges to come back. Why that is a design rule rather than an oversight is set out on /features/companion.
• It will not treat grief. MemoriesBox is a recording service. It is not therapy and we make no claim that it helps anybody heal.
• It will not beat a good printed book on the shelf. StoryWorth's hardcover, included from $59 a year, is better than ours today. Ours starts at $139 and is a separate purchase.

Where this sits against everything else

Of the eleven life-story services we compared in August 2026, none clones the storyteller's real voice and none lets the family ask a new question later. Two AI-legacy products do: Avataari, a Melbourne company that does not disclose its hosting region, and Afterlife AI, operated from New South Wales. MemoriesBox is the only one of them hosted in the European Union. The eleven, with prices verified on 17 August 2026, are on /compare.

Being fair about the rest: HeritageWhisper at $39 a year also runs a live adaptive interview and is much cheaper than we are. Remento at $99 includes a hardcover book. Storii phones a landline, still the best answer for someone who does not use a computer. Meminto is hosted in Germany. We are neither the cheapest nor the most established, and we are at version 0.4.1, which is pre-1.0.

What it costs

Questions people ask

What does MemoriesBox actually do?

It records a person's life story as a long guided interview across many sessions, keeps their own voice rather than a transcript, and lets their family ask that archive questions later and hear the answer in that voice. It costs $139 a year and is hosted in the European Union.

How is MemoriesBox different from StoryWorth?

StoryWorth emails a question a week for a year and prints the written answers as a hardcover book, from $59. MemoriesBox runs a long interview instead of a weekly prompt, keeps the audio as the thing that matters, and can be asked questions afterwards. The full comparison is on /compare/memorybox-vs-storyworth.

Do I need the device to use MemoriesBox?

No. The subscription works today on an ordinary phone or computer. The $79 device ships in 2027 and is for people who do not use a smartphone at all.

Where are the recordings stored?

In the European Union. They do not leave Europe, and they can be exported to your own drive at any time. The named country, provider and sub-processors are on /trust/where-your-data-lives.

---

Related: What happens to your voice, and who can reach it · Every other way to keep a voice · The full comparison table · How MemoriesBox works, start to finish

When you have read enough, start a subscription. $139 a year, first 14 days free, cancel anytime.