Account details
If you create an account, we store basics like your name and email so you can sign in, verify your address, and reach you about your account.
Trust
Privacy Policy
Last updated: August 5, 2026
KarmaLens is the product; this site is served at karmalens.ai. This page describes, in plain language, what we may collect today, why we use it, and what we do not do.
What we may collect
Google sign-in
If you use Google sign-in—for example by choosing Continue with Google on the login or signup page—the sign-in flow may share limited profile information with KarmaLens, such as your name, email address, and a stable identifier Google uses for your account, so we can create or sign you in. KarmaLens does not receive your Google password. Google sign-in is optional; you can use email and password on the same screens whenever both are shown.
KarmaLens does not intentionally store OAuth access tokens after sign-in. Tokens are used only to complete the exchange with Google; afterward your normal KarmaLens session applies. As stated elsewhere here, we do not sell your personal information.
Dilemmas and analyses
When you run an analysis, we process the dilemma text you submit and keep the structured result in your signed-in history (for example verdict, classification, scores, verse or teaching type, and share-friendly lines), so you can return to it later.
To produce a result, the dilemma text is sent to our language-model provider for that single analysis; we do not retain provider prompts or model traces in your history.
If you talk to your Sarathi about a reading, those messages are stored with that reading in your signed-in history so the conversation can pick up where it left off; deleting the reading deletes its conversation. Like the dilemma itself, conversation messages are sent to our language-model provider only to generate each reply.
Your first reading before signup
You can run your first reading without an account. The dilemma you submit is processed to generate that reading, and the result is held temporarily against your browser session so you can claim it when you sign up — nothing retyped, nothing lost.
Anonymous readings you never claim are not kept as account data. If a dilemma signals a crisis, we show the safety guidance immediately and store nothing.
Feedback you send
Optional feedback you choose to submit (such as usefulness signals, short tags, or a brief comment) is stored in line with the product’s feedback design—intended to improve quality, not to replace human judgment on every submission.
Technical logs
Like most sites, we keep routine technical records needed to run the service safely: for example errors, rate limits, and security signals. These are for reliability and abuse prevention, not for resale.
How we use this information
- To deliver analysis results and show them in your session.
- To maintain your account and saved history on the product.
- To improve clarity, safety, and usefulness over time.
- To detect misuse, protect accounts, and keep the service stable.
- To respond when you contact support.
What we do not do
No selling your personal story
We do not sell your personal information as a product category.
No hidden “always-on human review”
Feedback may be reviewed in aggregate or when you reach out—but the app does not promise that a person reads every submission in real time unless we say so plainly elsewhere.
Sign-in and payments
On karmalens.ai, Google sign-in is optional—you can use email and password on the same login and signup screens.
Paid checkout and billing are live; one-time purchases (and legacy subscriptions) are processed securely by Razorpay, our payment provider. The pricing and billing screens describe what applies.
Email categories
We send two kinds of email. Both come from [email protected].
Account-related (transactional) emails are tied to your use of the product: sign-up verification, password reset, payment receipts, security alerts. These are sent when the corresponding event occurs and cannot be turned off while you have an active account.
Optional (marketing) emails include the weekly newsletter, product updates, re-engagement reminders, and occasional discount or promotion notices. Each of these channels can be turned off individually on your email preferences page, or by clicking the one-click unsubscribe link in the footer of any marketing email you receive.
Turning off all marketing channels does not affect transactional emails.
Cadence: the weekly newsletter goes out on Monday mornings to every opted-in account. Paid members also receive a Saturday-morning email with personalized suggestions drawn from anonymized aggregate patterns — never your own past dilemmas. A separate reminder is sent at most once when an account has been inactive for fourteen days, and never again within the same inactive stretch.
Tracking: when you open or click a link in one of our marketing emails, our sending provider (Brevo) reports the event back to us so we can measure overall engagement and stop sending to addresses that consistently bounce or report spam. We do not use that data to profile individual readers and we do not sell it.
Public sharing (gallery)
You can request to share one of your own analysis results to the public KarmaLens gallery at karmalens.ai/gallery/. Sharing is opt-in per card; nothing is shared by default.
What gets published. The card — the verdict line, the share quote, the reflective question, and a theme label — plus a short, anonymized restatement of your dilemma. Before anything goes public, your dilemma is automatically rewritten into one tight, share-worthy line with every identifying detail removed: no names, employers, locations, or unique specifics. That rewritten line is the only form of your question that ever appears publicly. Your original dilemma text is never made public — it stays in your private history.
Identity is anonymous. Cards do not show your name, your username, your email, or anything else identifying. They appear as standalone reflections.
Moderation: AI advises, the moderation team decides. Every submission first runs through an automated pre-screen for personal-name leakage, abusive content, medical/crisis content, and topic fit. If the pre-screen declines a card, you are emailed with the reason and can re-share a different version. If the pre-screen passes, the moderation team explicitly approves the card before it goes live. The rewritten public line is independently checked for leftover identifying detail before it can be published, and the moderation team can edit it before approving the card.
Removal. You can request removal of one of your gallery cards at any time by emailing [email protected]; we will remove it within twenty-four hours. The moderation team can also remove a card after publication if necessary; in that case you receive an email noting the reason.
Discoverability. Approved cards are indexable by search engines so the gallery can carry reflective writing to people looking for it. We do not place tracking pixels on public card pages.
Personal share pages
Separately from the gallery, tapping Share on one of your readings creates a personal public card page. Sharing is opt-in per reading; nothing is public unless you tap share.
What appears. Only share-safe content: the shareable insight quote and the verse (quoted verbatim). Your dilemma text is never on the page. The page does not show your name, username, or email.
Not search-indexed. Personal share pages tell search engines not to index them — they are meant for the people you send the link to, not for search results. The curated gallery remains the public, discoverable surface.
Removal. You can remove the public page at any time from the reading itself; the link then stops working. Crisis readings can never be shared.
Cookies and similar storage
Essential storage keeps the product working: a sign-in session cookie, a security (CSRF) token, and a small browser setting that remembers your light/dark theme. These are always on while you use the site and need no consent — without them the site cannot function.
Advertising and measurement cookies come from Google Ads (gtag), which we use to measure how our ads perform. These stay off by default for everyone: until you press Accept in the cookie banner, the Google tag runs in cookieless mode — no ad-related cookie is set and nothing that identifies you is stored; Google receives only anonymous, aggregate signals. Pressing Accept allows the measurement cookies; declining keeps everything cookieless, and you can change your mind at any time.
You can change your choice whenever you like from the Cookie settings link in the footer, or by clearing site data in your browser. We do not place tracking pixels on public gallery card pages.
Sensitive dilemmas
KarmaLens offers reflective ethical guidance—not legal, medical, financial, or mental-health advice, and not an emergency service. Please avoid typing details you would not be comfortable storing on an account-backed tool, especially highly sensitive identifiers or third-party secrets.
Questions, corrections, and data requests
For privacy questions or to discuss your account data, email [email protected]. Self-serve export and automated account deletion are not promised on this page yet; we will describe them here when they ship. Until then: your dilemmas and saved results stay in your account history until you delete them (single rows or your whole history, from the dashboard) or delete your account — we do not auto-expire them, and we do not use them to train models.
When you delete your account we erase your data; we keep only a one-way, unreadable fingerprint of the email for about 30 days so deleting and re-creating an account can't be used to reset free quota. It cannot be turned back into your email and is deleted automatically.
Future changes
As KarmaLens grows—possible additions include new sign-in options or richer analytics—we will refresh this policy so the site stays honest about what is collected and why.