Privacy Policy

Last updated: August 25, 2026.

This privacy policy explains how MemoirWell collects, uses, shares, transfers and retains personal data.

Data controller

MemoirWell is operated by an individual autónomo (sole trader) established in Spain (EU): Danila Pryadko, NIE Z1106449Q, registered address C/ Alicante 4, PBJ 2, Madrid, Spain. As an EU controller we process personal data under the EU General Data Protection Regulation (GDPR) and the Spanish Ley Orgánica 3/2018 (LOPDGDD). For any privacy question or data request, contact hello@memoirwell.com.

Information we collect

MemoirWell collects account details, the memories and voice recordings you submit, any writing samples you provide for your voice profile, generated chapters, editorial outlines and fact-check notes, consent records, billing records, support messages, and the collaboration data you create — such as the email addresses you invite, project memberships, and share grants. We also keep technical information needed to operate and secure the service.

How we use information

We use this information to authenticate you, save your private memoir project, transcribe recordings, match your writing voice, generate requested drafts and editorial outlines, prepare fact-check questions, enable the collaboration you set up (sending invitations and giving the people you choose access to your project), provide support, process payments, prevent abuse, meet legal obligations, and improve reliability.

Optional product analytics

Product analytics starts off. We initialize PostHog and record events only after you choose Allow analytics. We send only a finite set of product events with coarse first-touch source, medium, campaign, content, and referrer categories. We do not send memoir content, recordings, chapter text, names, email addresses, search terms, raw URLs, or raw referrers. First-touch attribution expires after 180 days and is not renewed by activity. Declining or withdrawing permission stops capture and removes the saved first-touch state; analytics settings remain available on every page.

Legal basis for processing

In short: we handle your data to run the service, with your permission, and to meet the law.

We process your personal data on the following GDPR legal bases: performance of our contract with you (Article 6(1)(b)) to provide the memoir studio, generate chapters, keep your archive and manage payments; your consent (Article 6(1)(a)) for optional marketing emails and for processing the memory content you choose to submit; explicit consent (Article 9(2)(a)) where your memories include special-category data such as health, religion, politics, ethnicity, sex life or sexual orientation; our legitimate interests (Article 6(1)(f)) in securing the service, preventing abuse and improving reliability; and legal obligations (Article 6(1)(c)) for tax, accounting and compliance records. You can withdraw consent at any time.

AI processing

In short: I only read your memory to write your draft. We never train AI on your stories.

When you ask MemoirWell to create or revise a chapter, build an editorial outline, prepare fact-check questions, or match your voice from a writing sample, your submitted memory and text are processed by our primary Anthropic-compatible AI provider for the requested prose work. DeepInfra independently processes the source and candidate text needed for safety, rejected-claim, and content-retention checks with a pinned DeepSeek V4-Pro model, including when the primary prose provider is available. When the primary prose service has an eligible availability failure, eligible text-generation requests may instead be processed by DeepInfra using a separately pinned DeepSeek V4-Flash model. OpenAI is used only when you ask us to transcribe speech, with Groq available as an additional speech-to-text provider. These services receive the content needed to produce the result you requested or to verify that result before it is accepted. AI-generated chapters are drafts: you remain the author and are responsible for reviewing the facts, names, and wording before you rely on, share, or export them. MemoirWell does not use your memories to train AI models. DeepInfra states in its published privacy policy that normal inference inputs and outputs are not stored to disk and are not used to train models; each provider's handling is governed by its published terms and privacy notice. MemoirWell does not make solely automated decisions that produce legal or similarly significant effects about you. Your drafts and source memories stay in your private project — visible only to you and anyone you invite to collaborate — until you delete them.

Service providers we use

In short: a few trusted services help run MemoirWell, and they only see what they need to.

We share the minimum data needed with service providers: an Anthropic-compatible AI provider as our primary prose processor; DeepInfra as the independent processor for safety, rejected-claim, and content-retention checks and as the availability fallback for eligible text-generation requests; OpenAI for speech-to-text only; Groq as an additional speech-to-text provider; Polar (our Merchant of Record) for payments and tax handling; Resend to send account and collaboration-invitation emails; PostHog (EU region) only after optional analytics permission; and Hawk for error monitoring. When you invite a collaborator we share only the invitation email address needed to deliver that invite. We do not receive or store full card numbers. Each provider processes data under its own terms and privacy notice, and MemoirWell sends data only for the features needed to operate the service or requested by you. Where required, we use contractual safeguards for these service-provider relationships.

International data transfers

In short: some of this happens on servers abroad, always under EU-approved safeguards.

To turn speech into text and to generate chapter drafts, your memory text and audio may be processed by AI and speech-to-text providers located in the United States or other countries outside the EEA. These transfers rely on European Commission Standard Contractual Clauses (SCCs), data-processing agreements (DPAs), and, where available, adequacy mechanisms such as the EU-US Data Privacy Framework. We share only the minimum data needed and never use your memories to train AI models.

Your data-subject rights

Under the GDPR and LOPDGDD you have the right to access your data, to rectify inaccurate data, to erase your data, to data portability (export), to restrict processing, and to object to processing. You can exercise most of these directly in the app — export your stories, delete a session's voice recording, or delete your whole account at any time — and you can request erasure of specific chapters or transcripts by emailing hello@memoirwell.com. You also have the right to lodge a complaint with the Spanish data-protection authority, the Agencia Española de Protección de Datos (AEPD, www.aepd.es).

Your control and rights

You can export all of your stories at any time (your right to data portability under GDPR Article 20), delete any session's voice recording, revoke any invitation or share grant you have created, and delete your entire account in one click (your right to erasure under GDPR Article 17). To erase specific chapters or transcripts, email hello@memoirwell.com and we will remove them. Deleting your account removes your memories, generated chapters, and the collaboration access you set up from our active systems, except where limited billing, security or legal records must be retained. Any voice recording is used only to transcribe your session; it is never used to build a voice clone, and your voice profile is shaped only from the writing sample you choose to provide.

How long we keep your data

In short: we keep your stories until you delete them or close your account.

We keep your recordings, transcripts and chapters until you delete your recordings, request erasure of specific chapters or transcripts, or close your account. You can delete any session's voice recording in the app and request erasure of specific chapters or transcripts by email. When you close your account we erase your memoir content within 30 days; encrypted database and private-file backups are retained for 14 days; error and log events are PII-scrubbed and kept for about 90 days; and limited billing, tax, fraud-prevention or legal records may be kept where required by law.

Read the full Data Retention policy

Contact

For privacy questions or data requests, contact hello@memoirwell.com.