ABSOLUTE MEMORY
Your library and your data
Absolute Memory is a beta reading and knowledge app operated by Ash Pillai. This page describes how the app handles your information. Contact sakshay797@gmail.com with privacy questions or requests.
What the app stores
When you sign in with Google, Absolute Memory receives your name, email address, profile picture and account identifier. It stores the links, article text, notes, uploaded files, highlights, reading progress and conversations you save, together with generated summaries, tags, connections and other knowledge features.
Session cookies keep you signed in. Your browser also stores appearance and reading preferences. Operational records include usage counts, processing status, model and token usage, request duration and errors. The Chrome extension saves the page you choose and uses an account-linked capture token.
The Chrome extension
When you explicitly save a page, the extension reads its URL and title. Version 0.2.0 and newer also read available main document content. It removes scripts, forms and common navigation elements before sending a bounded HTML capture to Absolute Memory for article extraction. If Chrome cannot capture the page, Absolute Memory can attempt server retrieval. Version 0.4.0 also lets you explicitly save selected text as a quote. It sends that selection, its source, and any comment you add when saving. It does not collect browser cookies, continuously track browsing, or record your screen or keystrokes.
In the side-panel preview, a toolbar click can display a thumbnail from an image already loaded in the active page. That thumbnail stays in the extension.
Local extension storage holds an account-linked capture token, the app address and up to ten recent saves, including their titles, URLs, timestamps and preparation status. A bounded captured-page payload is retained for failed saves so a retry can survive a browser restart; successful saves discard that payload. A short-lived pairing verifier is kept in session storage. Disconnecting revokes the token and clears the recent-save list; it does not delete articles already in your account. The existing store release connects to relamai.com. The Absolute Memory update also supports app.absolutememory.co; existing connections remain available during the transition.
Limited use of extension data
Absolute Memory’s use of information received through the Chrome extension follows the Chrome Web Store User Data Policy, including its Limited Use requirements. We use this information to provide and improve the described saving, reading and knowledge features. We do not sell it, use it for personalized advertising, or use it to assess creditworthiness or lending eligibility.
Information is transferred to service providers as needed to deliver these features, or when necessary for security or legal compliance. Human access to extension user data is limited to specific user consent, security investigations, legal obligations, or aggregated and anonymized internal operations.
How AI processing works
When AI processing is enabled, saved text is processed automatically for organization and search. DeepSeek receives text and context for summaries, extracted concepts and ordinary reading or library questions. OpenAI receives text for search embeddings and, when you use the relevant features, images or PDF pages for extraction, text for narration, and questions for web search. These providers process requests under their own policies.
The optional web-search feature can send your question and relevant context to OpenAI. Saved-source answers and web results are identified separately. Your library is not a public collection, but the portions needed to provide these features are sent to the service providers.
For video imports, Supadata receives the selected public video URL to retrieve captions or generate a transcript. For podcast and audio imports, AssemblyAI receives the episode audio URL or uploaded recording when speech transcription is needed. Published podcast transcripts are retrieved from the feed’s source. Large uploaded audio files are temporarily stored in a private Cloudflare R2 bucket for validation and provider transfer. They are removed after transfer, with background cleanup and a one-day bucket expiry as safeguards. Temporary local audio bytes are removed after provider upload, or by background cleanup after one day; transcripts remain in your library. We request deletion of completed AssemblyAI transcripts through a retryable cleanup queue. Provider copies remain subject to their retention policies.
Hosting and source websites
The hosted beta uses Vercel for the website and application, with server processing in Ohio, Neon PostgreSQL in Ohio for storage, and Microsoft Azure in North Central US for scheduled background processing. Information may therefore be processed outside your country, including by AI providers. Google handles Google sign-in. Original article images may load from their source websites, which can receive normal browser request information.
Your controls and retention
You can export your library, delete individual items, revoke capture access, manage sessions and delete your account in Settings. Deleting an item removes its stored content and derived knowledge from the active database. Deleting your account removes its associated active application records.
We retain your saved library while your account exists. Database recovery copies and manual backups may retain earlier data separately until they expire or are removed. Provider records are governed by those providers’ retention practices. Contact us if you need help with access, correction or deletion, including backup copies.
Optional analytics and payments
If enabled and you opt in, PostHog receives a random browser identifier and predefined events such as a pricing-page visit or an upgrade click. We do not include saved URLs, article text, notes, chat messages or your email in these events. Session recording and automatic click capture are disabled. Use Privacy choices to change your preference; declining does not affect your account.
Paid subscription checkout is currently disabled. When enabled, Stripe handles subscription checkout and payment details. Absolute Memory stores billing customer and subscription identifiers, status, plan and billing period. Stripe may retain payment records under its own policies and legal obligations.
Updates
This page will be updated when the service’s data handling changes. This version was updated on 7 November 2026.