Privacy policy

Last updated: 2026-10-18 · Version: 2026-10-18.1

Pick Pick Go Inc. is the legal owner and operator of MemoSeek and the data controller responsible for your personal data. This policy applies to the MemoSeek app and this website.

Who we are

Legal name: Pick Pick Go Inc.

Postal address: 2315 E Olive St, APT 2B, Arlington Heights, Illinois 60004-5142, United States

Privacy contact: support@memoseek.ai

Sign-in

You can use email, or Sign in with Apple or Google where available. We receive the provider account identifier and the name/email you authorize. We do not receive your Apple or Google password; Apple’s private relay email is supported.

Scope

This policy covers the MemoSeek app, the servers it talks to, and this website. It does not cover other apps MemoSeek opens for you, such as Phone or WhatsApp, or services of Apple and other companies, which have their own policies.

What we collect

  • Account: your email address, your password (stored only as an Argon2id hash, never as text), your display name, language and dialect, timezone, and whether your email address is verified.
  • What you save: reminders (including the words you said or typed to create them, their time and timezone) and saved information such as places, numbers, warranties, purchases and facts.
  • Photos and documents you choose to add: the image, kept in private storage, and the details AI reads from it, such as a title, a summary and fields like dates and amounts. For an identity document this can include the document type, holder name, document number, nationality, issuing country, date of birth, and issue and expiry dates. Location data is removed from photos on your phone before upload.
  • Lists, expenses and people you enter: item text, notes, amounts, dates and names/phone numbers you choose to save; shared-item membership and activity; notification delivery identifiers where enabled. These are distinct from the phone address book, which stays on your device.
  • Calendar integration: connection data and authorization if you choose to link a calendar.
  • What you say or type: so that MemoSeek can understand it. MemoSeek does not store recordings of your voice (see "How your voice is handled").
  • Contacts: if you ask MemoSeek to find a person, the app reads names, organisations and phone numbers from the address book on your phone. Your contacts are not uploaded to our servers, and which contact a reminder is linked to is kept only in encrypted storage on your phone.
  • Device and sessions: an identifier for this installation of the app, the platform, app version and timezone, and your sign-in sessions. We also record where your account was created and where you use it: app or web, the device category (phone, tablet or computer), operating system, browser family and sign-in method — never the device model, serial number, an advertising identifier or your location — to understand how MemoSeek is used and to offer you the right update.
  • Usage and allowance accounting: how much of your AI allowance you have used, and the size and cost of AI requests. These records do not contain what you said or saved.
  • Support emails: if you write to us, your message and email address.
  • Emails we send you: verification codes and similar messages are queued with your email address; their content is stored encrypted and removed once sent.

Optional parking location remains on your device. We do not use advertising identifiers or advertising-tracking SDKs. Limited operational diagnostics support reliability and abuse prevention.

How your voice is handled

The microphone is only used while the app is open and in use, and the voice session stops when you leave the app. The microphone may open by itself when you open the Home screen, so you can speak hands-free, once you have allowed AI processing.

Depending on the voice mode your account uses:

  • Apple's speech recognition on your iPhone turns your speech into text. Apple may process that speech on its servers, under Apple's terms. The text is then sent to MemoSeek.
  • On MemoSeek Web, your browser's own speech recognition turns your speech into text: Chrome and Edge send the audio to Google, and Safari to Apple, under their terms. The text is then sent to MemoSeek.
  • For some accounts, your voice audio is streamed through MemoSeek's servers to Google's Gemini API so that it can be understood directly.
  • MemoSeek's code can also use OpenAI's realtime voice service; if it is enabled for your account, your voice audio goes to OpenAI.
  • When its transcription route is selected, Deepgram receives microphone audio and language settings to turn speech into text. The text is then handled by the selected understanding provider. Deepgram does not generate spoken replies.
  • When its transcription route is selected, and only if you agreed to the provider list that names it, AssemblyAI receives microphone audio and language settings to turn speech into text. The text is then handled by the selected understanding provider. AssemblyAI does not generate spoken replies.
  • MemoSeek does not store raw recordings of your voice. What you asked it to save is stored as text.
  • Most spoken replies are short pre-recorded clips shared by everyone. Other replies may be generated through the selected speech service. ElevenLabs produced the bundled generic recordings; playing them sends it no personal data.

Sensitive content

What you add may include sensitive information, document numbers, date of birth or nationality. Send only what you need and hide unnecessary fields before uploading an image. Reading identity documents is not identity verification or an end-to-end encrypted document vault. Extracted fields are stored on our servers if you ask to save them; we do not represent them as device-only.

Where it comes from

From you (what you type, say, photograph and choose); from your phone (the installation identifier, app version and timezone; contacts stay on the phone); from AI providers (the result of understanding your words or reading your photo); and from our own systems (sessions, usage accounting and security records).

Why we use it

  • To create your account, sign you in and verify your email address.
  • To understand what you say or type, and to save, show, search and remind you of what you asked.
  • To read the photos you add and help you find them.
  • To speak replies to you.
  • To keep your AI allowance and our costs within limits.
  • To keep the service secure and prevent abuse.
  • To answer your support requests.

We do not sell your data, use it for advertising, or build advertising profiles from it.

Legal bases

Where the law requires a legal basis, we rely on: contract, for providing the service you asked for; consent, for AI processing by third-party providers, which you can withdraw in the app at any time; and legitimate interests, for security and abuse prevention.

AI processing

MemoSeek uses AI providers to understand requests, read photos and generate speech. The app asks your permission before first use. If you choose "Not now", MemoSeek works in a limited mode: you can browse, edit and delete what you have saved, but voice, typed requests and photo reading are off. You can change your choice at any time in Settings → Privacy & Data → AI processing.

  • Google (Gemini API): may receive what you said or typed, with relevant recent conversation context and the relevant details of your saved items; your voice audio, for accounts that use the Gemini voice mode; photos you ask it to read; and text to be spoken, which may include your display name.
  • OpenAI (API): may receive what you said or typed, photos you ask it to read, text to be spoken (such as a welcome with your name, or a reminder's text), and your voice audio if OpenAI's voice mode is enabled for your account.
  • Anthropic (Claude API): may receive the text of your request and the relevant conversation, reminder or list context needed to understand it.
  • Deepgram: receives microphone audio and language settings for speech-to-text when its transcription route is selected. The resulting text is then handled by the selected understanding provider.
  • AssemblyAI: receives microphone audio and language settings for speech-to-text when its transcription route is selected, only for people who agreed to the provider list that names it. The resulting text is then handled by the selected understanding provider.
  • ElevenLabs: may receive microphone audio after you press Send for speech transcription when its route is selected and you have authorized that purpose, and reply text for speech generation. Playing the bundled generic clips sends it no requests or microphone audio.

Each provider handles data under its own terms. We do not claim that these providers never retain data or never use it for training; that is governed by their published terms. More on AI and your data and Subprocessors.

Sharing

We share data only with service providers that process it for us to run MemoSeek (listed on Subprocessors), and where the law requires us to disclose it. We do not sell personal data, share it for cross-context behavioural advertising, or allow tracking.

MemoSeek never sends a message or places a call for you: the Call and WhatsApp buttons open those apps, and you act there.

International transfers

Our servers and database are hosted by Amazon Web Services in the United States (region us-east-2). AI providers may process data in other countries, as their terms describe. If you use MemoSeek from outside the United States, your data is transferred to and processed in the United States.

Retention and deletion

  • Raw microphone recordings: MemoSeek does not persist them. Audio is processed in memory for the requested turn; external provider retention is governed by the provider’s terms.
  • Saved reminders, lists, people, expenses and information, including selected photos/documents: kept while your account exists or until you delete them. Deleted-record synchronization cleanup follows the policy below.
  • Unfinished workflows: kept until you complete or cancel them. For expired drafts no longer attached to an open workflow, previous utterances and side context are cleared at expiry; structured draft state is cleared after the seven-day offline-commit grace period, and operation receipts are swept after 30 days.
  • Generated reminder audio: removed after the reminder is no longer active or a day after its scheduled time, by the audio cleanup process.
  • Support messages: kept to resolve your request and meet related legal obligations. You can ask support to erase a closed conversation, subject to those obligations.
  • Limited content-free credit, usage and security records: retained for accounting, abuse prevention and security obligations, including after account closure. These can retain a closed-account identifier; they are not represented as fully anonymous.
  • Provider-held data follows the provider’s applicable terms. Deleting your account does not promise immediate erasure of independently retained provider security records.

Deleting an item and deleting an account are different. A deleted item disappears from your active lists immediately; synchronization records marked as deleted are swept after 30 days. Shared records or live references may keep a record for other participants, with private owner context removed by the retention process. Account deletion does not wait for that window: it revokes sessions and erases private database content in the deletion transaction. Photo, document and generated audio objects are queued for removal from storage afterwards. Content already shared with other people may remain in their records; remove sensitive shared content before closing your account. Limited content-free usage, security and accounting records can remain under a closed account identifier.

Account deletion instructions: Delete your account.

Operational diagnostics

Operational records include connection events, request timing, success/error status and credit usage linked to an account/install identifier. They are not advertising tracking. Standard cleanup windows are 30 days for events, 7 days for detailed diagnostics when enabled for a support case, and 180 days for incidents. Microphone recordings are not uploaded for diagnostics. Content diagnostics are limited to enabled support/test modes and may include request text and its result.

Security

Connections use HTTPS, passwords are stored as Argon2id hashes, and photos are kept in private storage. On your phone, your items are kept in an encrypted file and your sign-in details in the iOS Keychain. MemoSeek is not end-to-end encrypted: our servers can read what you save, so that the service can work. More on the Security page.

Your rights and how to use them

Depending on where you live, you may have the right to access, correct or delete your data or receive a copy of it, to object to or restrict some processing, and to withdraw your consent.

In the app: Settings → Privacy & Data, with "Download my data", "Delete my account" and "AI processing". You can edit or delete reminders and saved items directly. For anything else, write to support@memoseek.ai from the address on your account; we may need to confirm it is you. You can also complain to the data protection authority where you live.

The detail is on Privacy choices.

Children

MemoSeek is for people aged 13 or older. You must be at least 16 where local law requires that age for consent to digital services, including the relevant European countries, or meet any higher applicable minimum. If you are below the legal age of majority, use the service only with a parent or guardian’s permission. MemoSeek is not directed to children below these limits.

This website

The public pages use no advertising-tracking cookies or marketing analytics. Cloudflare protects and delivers the website and may process IP addresses/request headers and use security cookies where needed. Cloudflare Email Routing forwards support messages to the company-managed mailbox.

Contact

For privacy questions and requests: support@memoseek.ai.

Changes to this policy

This policy is versioned; the version and the date it was last updated are shown at the top. A material change is published here with its effective date.