Hypeless ("we", "us") is a free educational website run by Kriplani and Kothari, Jaipur, Rajasthan, India. This page explains what information we collect and why. We wrote it to follow India's Digital Personal Data Protection Act, 2023 and the Information Technology Act, 2000 and its rules.

The legal operator of Hypeless, responsible for your data, is Kriplani. Our address and phone number are on our Contact page.

The short version

You don't need an account. The news, the Tip Checker, SEBI check and Scams work without signing in.
If you choose to sign in with Google or Apple (to save stories or follow companies), we keep your email address (with Apple, this can be Apple's private relay address), your name if the sign-in service shares it, plus the stories you save and the companies on your watchlist.
We don't use advertising or outside analytics tools. We count page views and how often features are used ourselves, without personal data. One small first-party cookie (hl_seen) may mark that this browser has visited before, so we can count return visits (see "Counting visits").
If you joined the Pro waitlist before Pro opened, we keep that email only as described below. The waitlist is closed.
The Tip Checker: when you press Check this post (signed in or not), our rules check the post on your own device, and the post is also sent to our server and, through our AI gateway Mesh API, to DeepSeek's AI (China), or to Google's Gemini AI as a backup if DeepSeek fails, for the AI check. If a Pro member presses Try again, the post goes to both of them, and to Claude from Anthropic if they disagree. A screenshot you upload is sent the same way. Unless you untick the box, we keep the post you typed or pasted for up to 8 days so a second AI can re-check our answer (see "A second AI re-checks our answers"); a screenshot isn't kept. We keep it longer only if you send it with a mistake report or share it (see "Sending us a post with a mistake report"), and a one-way code spots repeats (see "When the same message is checked again"). Nothing is sent until you press the button. See "The Tip Checker's AI check" and "Which AI companies we use".
Ask Hypeless needs an account. Your questions go to the same AI services, and your chats are saved to your account until you delete them (see "Ask Hypeless and your saved chats").
Payments for Pro are handled by Cashfree. We never receive your full card, UPI or bank details, and we don't save the mobile number you type at checkout. We keep only payment records (see "Payments for Pro" below).
If you make API keys or buy API credit, we keep the key names, a scrambled copy of each key, daily call counts and a record of your credit (see "API keys and credit").
If you send us feedback through the form, we keep your message and an email address to reply to (the one you type, or your account email if you are signed in) for up to 12 months (see "Sending us feedback").
We don't sell personal data, and we don't share it except as explained below.

What happens when you visit

Our website is hosted by a hosting company. Like almost every website, the hosting company's computers may automatically keep basic technical records (such as your IP address, browser type and the time of your visit). These are used to keep the site working and secure. We don't use them to identify you or build a profile of you.

The Tip Checker's AI check ("Check this post") and the screenshot checker (updated 30 October 2026)

When you press Check this post, our software on your device first finds any SEBI registration number and company names in the post, and then the post is sent for the AI check, which finds the red flags. This happens whether you are signed in or not (signed out: up to 3 AI checks a day per device; a free account 5; Pro 50). You can also upload a screenshot of a post, which is sent the same way. Nothing is sent until you press the button or upload a screenshot. If you share a post or a screenshot to Hypeless from another app (Android, with the app installed), that counts as pressing the button: it is checked straight away. While the page opens, the shared post waits for a moment in your phone's own storage and is then deleted. When a post is sent:

What is sent: the text of the post you pasted (or your screenshot), and a random code your browser keeps for this feature (see below). Like every request on the internet, your IP address also reaches our server.
Who receives it: our server; Mesh API, an AI gateway service (meshapi.ai, run by Fiesta Labs Inc., USA) that we pay for AI use and that passes the request on; DeepSeek, an AI model from China (our first choice); and, only if DeepSeek fails or is too slow, Google's Gemini AI (Google LLC, USA) as a backup. We send them only the post, a list of recent news headlines and the SEBI numbers and company names our software found. We don't send your IP address or the random code. Mesh API says that, with its default setting, which we use, the content of requests (what is sent and the answer) is not written to its logs or databases, that it keeps only billing details such as the model used, token numbers and cost, and that it doesn't use content to train AI (its privacy policy, checked 6 November 2026). It doesn't say where requests are processed, and the AI company that receives a request follows its own rules. DeepSeek's and Google's computers are outside India (DeepSeek's may be in China), and they handle the text under their own terms: DeepSeek's privacy policy, and Google's Gemini API terms and privacy policy. Anything you paste, upload or ask only ever goes to DeepSeek or Google, never to another AI company, with one exception: Claude from Anthropic on a Pro member's Try again (see above).
What we keep: unless you untick the box under the Check button, we keep the post you typed or pasted for up to 8 days so a second AI can re-check our answer (see "A second AI re-checks our answers"). We don't save the AI's written answer, and we don't write posts to our logs. We also record that an AI check happened and how many AI "tokens" it used (a number, to watch our costs), with nothing about you. To limit how many AI checks each internet connection can use per day, our server keeps a scrambled (hashed) form of your IP address and, if you are not signed in, of the random code (or, if you are signed in, of your account code), in memory only, until midnight India time. Every signed-in account has an AI budget in rupees (a small one for a free account, a bigger one for Pro): a rolling 5 hours and an India calendar month. To keep it the same on all our servers, our database saves one line for each AI call or web search you start: your account code, the feature name (for example "tipCheck"), what it cost in rupees and the time. Nothing about the post, the image, the question or the answer. Lines older than about five weeks are deleted. The usage meter on the Tip Checker and Ask pages shows you these numbers.
Screenshots: your browser first shrinks the image and removes hidden photo information (such as location). The image is then sent to our server and, through Mesh API, to DeepSeek (or Google's Gemini as a backup) only to read the post's text and check it. We don't save the image or its text, and we don't write them to any log. The copied text is shown back to you on your device.
Why: only to give you the AI check you asked for, and to keep the daily limits fair.
Please don't paste personal details (phone numbers, bank or UPI details, names of private people) into a post you check with AI. Screenshots can show other people's names, numbers or messages, so crop the image to just the post before uploading. Our AI is told to ignore names, phone numbers and app details, but it may still see them.

When you last opened Hypeless while signed in

If you are signed in, we save the date (only the date, never the time or the page) you last opened Hypeless on your account. Each visit replaces the old date. We use it only to count how many accounts are active. It goes when your account is deleted.

Counting visits (first-party, no personal data)

To know which pages are useful, our own server counts page views and how often features are used (for example "a Tip Checker check ran", "a share", "a SEBI lookup", or "a payment succeeded"). For each count we store only: what happened, the page address (without anything you typed or searched), the date and time, and the country, state or region and city your connection comes from (worked out by our hosting company from your connection, for example "IN", "RJ" and "Jaipur"; this is approximate and can be wrong). We use it only to see in which parts of India and the world people read Hypeless. We never store your IP address, we never give you an analytics ID, we never send these counts to Google Analytics or similar tools (page speed numbers are measured separately, see below), and we never link counts to your account, so the counts can't tell us who you are. We also count when an AI app connects to our MCP server or uses one of its tools, with the app's kind only (for example "Claude"), never who or what was asked. Payment counts are only "a payment succeeded" — never the amount, card or UPI details (those stay in the payment record described above, if you buy Pro).

A few more labels on each count (added 21 October 2026). Next to each count we may also store a short label from a fixed list, never the raw text your browser sent: the kind of device (phone, tablet or computer) and the browser family (for example Chrome or Safari), both worked out from your browser's user-agent line, which we then throw away; for the first page of a visit, only the domain of the website you came from (for example google.com, never the full address or your search words) or "direct"; the language you picked on Hypeless (English, Hinglish or Hindi), read from the lang cookie you already have; and whether you used the website or our iPhone app. We also count a few steps and features (a payment started, an Ask answered and the power level used, why an Ask answer failed, which part of the site returned an error, a Quick news card saved or skipped), and for each AI call how long it took and whether it worked. These are all counts of how often something happens. They are never linked to you, your account, your IP address or anything you typed, and we do not use them to follow you from one website or visit to the next.

Page speed. Until 31 October 2026 we used Vercel Speed Insights (page load times, page address, device and browser type, country; no cookies, nothing about who you are). We have switched it off.

To count return visits (a browser that has been here before), we set one small first-party cookie called hl_seen with the value 1 for up to a year. It holds no name, email, account code or random ID — only that this browser has opened Hypeless before. We do not use it to follow you across other websites. You can delete it any time in your browser settings; then the next visit counts as new again. If your browser sends "Do Not Track" or "Global Privacy Control", we don't count you at all and we don't set that cookie.

Sometimes we may test a short question under a Tip Checker result, "Would you trust this tip?" (Yes, Not sure or No). If you answer, we count it the same way: the answer word, whether you were asked before or after seeing the result, the page and the day. Never who you are, never the post. Which group your browser is in is kept only on your device.

Breaking-news notifications (optional, no account)

On the news page and the home page you can press Notify me to get a short notification when three or more named news outlets report the same big story. Nothing is asked or stored until you press the button and allow notifications in your browser.

What we store: your browser's push subscription: an address at your browser's push service (Google, Apple, Mozilla or Microsoft) and two keys that let only your browser read our messages. We do not store your name, email, account, IP address or any device ID with it, and it is not linked to your account if you have one.
What we send: only the outlet, the headline and a link to the story page on our site, for example "Outlet: headline". At most 4 a day, and never between 11 pm and 7 am India time. The message is encrypted so the push service can't read it. We also keep a short record of which story we sent (outlet, headline, date) so nothing is sent twice; it is deleted after 7 days.
How to delete it: press Turn off on the news page or the home page and we delete your browser's subscription straight away. You can also block notifications for hypeless.in in your browser settings; we then delete the subscription the next time we try to send. Because we hold no name or email for it, we can't find a subscription for you by email, so Turn off or your browser settings is the way.

Your language choice

If you pick a language (English, Hinglish or Hindi), your browser keeps one small cookie called "lang" for up to a year, so the site stays in that language. It holds only the language ("en", "hinglish" or "hi"), nothing about you. We don't use it to track you, and you can delete it any time in your browser settings.

Signing in with Google or Apple (optional)

You can create a free account by signing in with Google. We never see your Google password. When you do:

What we keep: your email address, your name and the link to your Google profile picture (Google gives us these when you agree on its screen), when you joined, your answer to the age / parent question, and whether you have Pro. If you use them: the stories you save (headline, news outlet and link) and the companies on your watchlist.
Why: to sign you in, to keep your saved stories and watchlist, and (later) to know whether you have Pro. We don't send you marketing emails and we don't sell or share this data.
Who stores it: Supabase, the company that runs our database and sign-in, on servers in Mumbai, India. Google handles the sign-in itself under its own privacy policy.
Signing in with Apple instead: if you choose “Sign in with Apple”, we never see your Apple ID password. Apple gives us your email address, or, if you pick “Hide My Email”, a private relay address (ending in @privaterelay.appleid.com) that forwards to you. Apple shares your name only the first time you sign in, and only if you allow it. We keep the same things as for Google (except a profile picture, which Apple doesn't give), for the same reasons. Apple handles the sign-in under its own privacy policy.
Login cookies: signing in puts small cookies on your device that keep you signed in. They are needed for the account to work and are removed when you sign out. They are not used for tracking or advertising.

Your watchlist

When you are signed in, you can follow companies on your portfolio (your watchlist).

What we store: which companies you follow (a short code for each company from our own list, like "tcs"), linked to your account, and when you added each one. Nothing else.
Why: only to show you the latest headlines about those companies (and, with Pro, your daily brief). Following a company is not advice, and we don't use your watchlist for advertising or share it.
Who can see it: only you. Our database lets a signed-in person read and change only their own watchlist.
How to delete it: press Following next to a company on your watchlist to remove it straight away. To delete your whole account, including your watchlist, email support@hypeless.in.

Your portfolio on the news page (updated 22 October 2026)

On the news page you can paste or upload a list of the companies you hold ("My portfolio"), or type it in your own words, for example "HDFC Bank 50, tata motors, Reliance".

What happens to the text: your browser first removes quantities, prices, ISIN codes and column headings, so only company names are sent to our server. Our server matches them to the companies we know (every company listed on NSE or BSE and a few well-known private ones) and forgets the text at once. We never store the text, quantities or prices.
AI for the lines we cannot place: if you are signed in and a line is a nickname or has a typo (for example "HDFC"), our server sends only those names (never your account, email or anything else) through Mesh API to DeepSeek, or to Google's Gemini as a backup, and the answer is checked against our own list: the AI can only pick a company that exists there. Mesh API handles it like every AI request (see "Which AI companies we use"), and it counts towards your AI spending record. Signed out, this step is skipped and nothing goes to an AI.
What we keep: signed out – the list of company codes (like "hdfcbank") stays on your device only (in this browser's storage), and we never see it. Signed in – the same list of company codes is also saved to your account, in the same place as your watchlist (see "Your watchlist"), so your portfolio follows you to your other devices; when you sign in, the list on the device and the list on the account are joined. A free portfolio holds up to 2 companies, and Pro has no limit. If you paste more than a free portfolio holds, the extra company codes wait on your device only (in this browser's storage, never sent to us) so you do not have to type them again; Clear all or "Forget the waiting list" removes them. Nothing else is stored: no quantities, no prices, no file.
How to delete it: press Clear all in the portfolio box (it also clears the account copy while you are signed in), or remove companies one by one. To delete your whole account, email support@hypeless.in.
Company search: when you type in a company search box, the letters you type are sent to our server to find matching company names. They are not stored or linked to you.

The Pro daily brief

Pro members see a short AI-written brief on their watchlist page. To write it, our server sends DeepSeek's AI, through Mesh API (see "Which AI companies we use"), only the names of the companieson your watchlist and our own recent news headlines about them. We don't send your name, email, account or IP address, so the list of companies can't be linked to you. The brief is saved for the rest of the day under the list of companies (not under your name), so anyone who follows exactly the same companies that day sees the same brief. To keep the daily limit fair, our server keeps a scrambled (hashed) form of your account code in memory only, until midnight India time.

Market close summaries, weekly recaps and the notice banner

The "Market close" summaries, the weekly recaps and the occasional notice at the top of the page are written by AI once for everyone, from public headlines by our named outlets. No personal data is involved: nothing about you is sent to anyone to write them, and reading them sends us nothing. If you close the notice, your browser remembers that on your own device (in its local storage); it is never sent to us. The live market brief on the home page is also written once for everyone from public headlines only, by DeepSeek's AI through Mesh API.

Which AI companies we use

Claude (from Anthropic, USA, through Mesh API): used only when a Pro member presses Try again with 2 AIs in the Tip Checker and our other two AIs disagree on a serious warning sign. It then receives the post and the headlines used for the check, never your name or account. See Anthropic's privacy policy.
Mesh API (meshapi.ai, run by Fiesta Labs Inc., USA): the gateway every AI request goes through. It passes each request to the AI companies below. It says that, with its default setting, which we use, it doesn't store the content of requests and doesn't use it for training. See its privacy policy. AICredits (aicredits.in, India) is our backup gateway: if Mesh API fails, a request can go through it instead, and requests before 6 November 2026 went through it. It says it keeps request content for 30 days and then deletes it. See its privacy policy.
DeepSeek (an AI model from China; Mesh API may run it with DeepSeek in China or with another hosting company): our first choice for every AI feature. It receives what you paste, upload or ask when you use the Tip Checker's AI check, the screenshot checker or Ask Hypeless, and public news headlines (and, for the Pro daily brief, the names of companies, never who follows them) for the live market brief, the Pro daily brief, the results summary and the free "Quick read" story summaries. See DeepSeek's privacy policy.
Perplexity (USA): only the restricted web search described below.
Google (Gemini, Google LLC, USA): a backup only. If DeepSeek fails, gives an unusable answer or is too slow, the same request (the same post, screenshot, question or headlines) is sent once to Google's Gemini instead. See Google's Gemini API terms and privacy policy.

Ask Hypeless and your saved chats

If you ask about "my watchlist" or "my portfolio", our server reads the companies on your watchlist and picks headlines and filings about them. If you saved a portfolio on the news page ("Upload your portfolio"; it is kept on your device, and on your account too when you are signed in, see "Your portfolio on the news page"), the chat also sends the names of those companies (never quantities or prices) with that one question, so we can pick headlines about them; our server does not store them. The list itself is not sent to DeepSeek, Google or Perplexity, only those public items and your question. The paper clip button in the chat uses the portfolio import described below: the file is read in your browser, only company names are sent to our server, and quantities, prices and the file are never stored.

Ask Hypeless needs you to be signed in. When you press Ask, your question is sent to our server and, through Mesh API, to DeepSeek (or Google's Gemini as a backup, see above), together with public material our server picked for the question: recent headlines from our named outlets, titles of official company filings on BSE, NSE's public list of IPOs, our own market summaries and Learn articles, and short extracts from the restricted web search described below. The last few questions and answers of the same chat go too (so a follow-up question makes sense). Nothing is sent unless you press the button. We don't send your name, email, account or IP address to DeepSeek or Google.

We save your chats to your account (the "ask_chats" table): a title (your first question, cut short), your questions, our answers (with their sections) and the sources each answer cited (headlines, filings and so on), and when the chat was last used. Only you can see them. They don't expire: they stay until you delete a chat (the cross next to it on the Ask page) or your account (write to support@hypeless.in). There is a limit on how many chats are kept per account; past it, your oldest chat is deleted. We don't write your questions or answers to our logs and we never use them for advertising. We also record how many AI "tokens" each answer used (a number). For the limits, our server keeps a scrambled (hashed) form of your IP address, in memory only, until midnight India time, and your questions are paid from your AI budget in rupees, which our database records as one line per question (your account code, the feature name, the cost in rupees and the time; no text; deleted after about five weeks) so the budget is the same on all our servers. The answer is written in the language you picked on the site (the "lang" cookie); the same is true for the other AI features.

Rating an Ask Hypeless answer (thumbs up or down)

Under each answer you can press a thumbs up or thumbs down. We then save one row for your account (the "ask_feedback" table): your account code, the rating (up or down), the answer mode you used, the language of the answer and the time. We save the text of your question and our answer only if you tick the box "Share this question and answer to help improve Hypeless" before you press a thumb; the box is off unless you tick it. Our team reads shared questions and answers to improve the prompts and tests of Ask Hypeless, and ratings are counted. We never use them for advertising and never sell them. Only our team can read this table. You can ask us to delete your ratings and shared text at any time (support@hypeless.in); they are also deleted if your account is deleted.

Once a week our AI services (through our AI gateway, to DeepSeek, or Google as a backup) read the questions and answers people chose to share after rating an answer badly, along with counts of how our AI features did, to write a report for our team. It never includes your name or account.

Restricted web search (for the AI features)

When it helps (for example when our own headlines don't cover something, or you ask why something happened or what a word means), our server may run one web search, limited to the news outlets we already show, Reuters, Bloomberg and official websites (BSE, NSE, SEBI, RBI, PIB, the Ministry of Corporate Affairs, AMFI, NSDL, CDSL). For "what is..." questions Wikipedia is also allowed, only to explain a word. Forums, social media, blogs and stock-tip pages are never used. The search goes through Mesh API to Perplexity (Perplexity AI, Inc., USA), which runs the search under its own privacy policy (on perplexity.ai). What the search contains:

Ask Hypeless: the text of your question (never your name, email, account or IP address).
The Tip Checker's AI check and the screenshot checker: only the names of the companies in the post and words such as "results" or "order" chosen by our code. The post itself is never sent to the search service.
Quick read: the public headline of a story that came with no short summary from its outlet.

Our server then opens at most three of the result pages and keeps only a short extract (a few sentences) for the AI, which must name the outlet and link to the original. We keep each search and its results in our server's memory for up to 6 hours so the same search isn't paid for twice, and then they are gone. Mesh API handles the search request like every AI request. There are daily limits on searches.

Portfolio import (free up to 2 companies, more with Pro)

If you paste a list of companies on your watchlist page, our server matches the names to the companies we know and then forgets the list. We never store the pasted text, quantities or prices. Only the companies you confirm are added to your watchlist, exactly as if you pressed Follow.

The company list

Hypeless knows the companies listed on NSE and BSE and a few well-known private ones. The list comes from the exchanges' own public lists. Once a day our server checks NSE's public list for newly listed companies and saves their public details (name, NSE symbol, ISIN) in our database. This is company information only; it holds nothing about any visitor.

The news archive

We keep a copy of public headlines from our named outlets for about 13 months (headline, outlet, link, time, the feed's own short description and which companies it mentions), so company pages and My portfolio can show past coverage (at least the last 6 months). It holds no information about you or any visitor.

Pro watchlist alerts

Pro members can get alerts when a named news outlet reports on a company on their watchlist. An alert is only the outlet's name, the headline, the time and a link to the outlet.

What we store: which stories we alerted you about (headline, outlet, link, the company and the time), linked to your account, and whether you have read each one. Alerts are deleted automatically after 30 days.
Notifications (only if you turn them on): when you press Turn on alerts and allow notifications, we store your browser's push subscription: an address at your browser's push service (Google, Apple, Mozilla or Microsoft) and two keys that let only your browser read our messages. We also store your choices (notifications on or off, quiet at night). The push service delivers the notification; the message is encrypted so it can't read it.
How to delete it: press Turn off alerts on your watchlist page to delete every push subscription for your account straight away. To delete your alerts, email support@hypeless.in.

IPO tracker and results-day digest (Pro)

The IPO tracker shows IPO names and dates from NSE's public list and public headlines from our named outlets. Reading it sends us nothing about you.

Following an IPO: if you are a Pro member and press Get alerts on an IPO, we store only that IPO's id and name (for example "Vishal Nirmiti") and the time you pressed it, linked to your account, so the alerts job can tell you when a named outlet reports on it. Press Alerts on again to delete it straight away. It is never used for advertising or shared.
Results-day alerts: when named outlets report quarterly results for a company on your watchlist, you get one alert for that company and day. It is stored like the other alerts above (the title, the outlets' names, the company and the time) and deleted after 30 days. No new information about you is collected.
The plain-words results summary: written by AI (DeepSeek, through Mesh API) from public headlines only, once per company, day and language, and shared by everyone who sees it. Nothing about you is sent to write it.
"Quick read" (short AI summaries of up to 200 words on news cards): written ahead of time, as soon as a story appears in our feed (or the first time anyone opens one that has none yet). Our server sends only that story's public headline(s) and the outlets' own short summaries to DeepSeek's AI through Mesh API; the summaries are saved and shared by everyone. Nothing about you is sent, and we don't record who opens which summary.

The Pro waitlist (closed)

Before Pro opened, people could join a waitlist on our Pro page. For those people we keep the email address, the date they joined and which button they used, only to tell them that Pro is open. No new emails are collected. We don't share them, and you can ask us to remove yours at any time.

Extra AI uses (Pro). If you buy extra AI uses, we keep your account's balance of extra AI (in rupees), a record of each change (bought, refunded) with its order ID and time, and, for each AI call, whether it was paid from the extra balance. Nothing about what you asked is stored with it.

Payments for Pro and API credit

Who processes payments: Cashfree (Cashfree Payments India Private Limited), a payment company in India. You enter your card, UPI or bank details on Cashfree's secure payment page, not on ours.
What we never receive: your full card number, CVV, UPI PIN, net-banking password or bank account login. Cashfree collects and handles those under its own privacy policy, which applies to what it collects.
Your mobile number: Cashfree requires a mobile number for every payment, so we ask for it when you press "Get Pro". We send it to Cashfree, together with your account's email address and an account number (not your name), to create the payment. We don't save the mobile number, and we never use it to contact you.
What we store: one record per order: our order ID, Cashfree's payment ID, what was bought (the Pro plan, which can be early, monthly or yearly, and its number of days including any yearly bonus days, the API credit pack, or a pack of extra AI uses and how many AI rupees it adds), the amount and currency, its status (for example created, paid or expired), when it was made and paid, the date your Pro ends, whether it was a test payment, and the email address of the account it belongs to (Google, or Apple, which may be a private relay address).
Why: to switch Pro on for the right account, to show you your plan, to handle refunds, to answer your questions about a payment, and to keep the accounts and tax records the law requires. Cashfree also tells our server directly when a payment succeeds, so Pro switches on even if you close the page.
How long: payment records are kept for as long as Indian tax and accounting laws require us to (this can be several years), even if you delete your account. We delete them after that. We never use them for advertising, and we don't sell or share them except with Cashfree, and with our accountant or the authorities when the law requires it.

API keys and credit

Keys: if you make an API key on our API keys page, we keep its name, its first few characters, a scrambled copy (a hash, so even we can't read the key back), when it was made, last used and revoked. The full key is shown to you once and never stored.
Use: we count calls per key and per account per day (numbers only: how many calls, how many AI questions, how much credit they used). We don't keep what your app asked, except that an AI question is sent to our AI services to answer it, as for Ask Hypeless, and not saved.
Credit: we keep your credit balance and a record of every change: purchases (with the order ID), one line per day for calls, and any refund. We need it to charge calls correctly, to show you your balance, to handle refunds and to keep the accounts the law requires.
How long: keys, daily counts, your balance and its record of changes stay until you delete your account, and are deleted with it. The payment records of credit you bought are kept like other payment records (see above), as tax and accounting laws require.

Sending us feedback

The Send feedback form works with or without an account. When you press Send we save one row (the "site_feedback" table): the type you chose (bug, idea or other), your message (up to 1,000 characters), the page you were on before opening the form (only its address path, for example /news, never the part after a question mark), the language of the site and the time. We save an email address so we can write back to you: the one you type, or, if you are signed in, the email of your account (we do not save your account code). It is used only to reply. We do not save your IP address with it (the form only counts how often one connection sends messages, in memory, with a scrambled address, to stop spam). Our team reads the messages to fix problems and improve Hypeless. We never publish them, never use them for advertising and never sell them. Please do not type passwords, card details or other private numbers in the message. Messages are deleted after 12 months, or sooner if you ask us (support@hypeless.in, from the email address you gave).

Reporting a mistake

Every story, every Tip Checker result and the Corrections page have a Report a mistake button. It works with or without an account. When you press Send we save one row (the "error_reports" table): the page it is about (only its address path, for example /story/ab12), the kind of mistake you chose, your message (up to 1,000 characters), the language of the site and the time. We do not save your name, email, account or IP address with it (we only count how often one connection sends reports, in memory, with a scrambled address, to stop spam). Our team reads the reports to fix mistakes. When we fix one, we publish a dated correction saying what was wrong and what we changed, never who reported it, and every three months we publish the number of reports and corrections. Reports are deleted after 12 months.

Rating a story ("Not market news")

Under every story there is a Not market news button. It works with or without an account. When you press it we save one row (the "news_votes" table): the link of the story, its headline as the outlet published it, your vote, the language of the site and the time. We do not save your name, email, account or IP address with it (we only count how often one connection votes, in memory, with a scrambled address, to stop spam). Your phone or computer remembers which stories you voted on (in its own storage, never sent to us), so the button shows your vote. When several people say a story is not market news, it is taken off the site, and our AI reads the headlines people flagged so it can choose better stories in future. Votes are deleted after 90 days.

A second AI re-checks our answers

To catch our own mistakes, a second AI re-checks Tip Checker answers. If the box under the Check button stays ticked, we save the post you typed or pasted and the warning types it got, after removing the phone numbers, emails, payment IDs and bank numbers we can spot. This can miss things, so don't paste anything private. We save no name, account, device code or internet address with it. Through our AI gateway Mesh API, Google's Gemini reads the post without seeing our answer, and we compare the two. Our team can read the posts where the two AIs disagree, and those go to our AI services for a weekly report. Each post is deleted within 8 days, and it never appears on a public page from this. Untick the box and the post isn't kept at all. Screenshots aren't part of this.

When the same message is checked again

When an AI check finishes, we keep a one-way code made from the message's text, the warning types it got, how many times it has been checked and the last time. The message can't be read back from the code, and it isn't linked to you, your account, your device or your internet address. If exactly the same text is checked again, the AI is told the earlier warning types so it can look at them again, and the page can say the message has been checked before. Each code is deleted 30 days after it was last checked.

My past checks on the Tip Checker page are kept only in your own browser and never sent to us. Anyone using that browser can see them, and the Clear button removes them.

Sending us a post with a mistake report

Under a Tip Checker result you can report a mistake. Only if you tick the box to send the post you checked, we save it with your report (the "error_reports" table), along with the warning signs we showed. If you don't tick it, the post stays unsaved as described above. Before saving, we remove the phone numbers, email addresses and UPI IDs we can spot, but this can miss things, so don't send anything private. We don't save your name, account or IP address with it.

Why: to find what the checker gets wrong and improve its rules. Made-up versions of a post may become test cases. Rules learned from mistakes are added to the instructions our AI reads on every check, and are listed publicly on our track record page.
Our public scam messages page: if a post you send looks like a scam, it can be published on /scams/messages and on pages for companies it names. Our AI checks it first: it goes up by itself only if the AI is sure it's a scam and nothing in it points to a person, and otherwise it waits for our team. It isn't linked to you. Our team can take any message down: to have one removed, write to us.
Counting repeats: when anyone checks a post, our server works out a one-way code from its words and adds one to a counter if that code matches a message already on that page. If it matches nothing, nothing is saved.
Who sees it: our team, and once a week our AI services (through our AI gateway, to DeepSeek in China, or Google's Gemini as a backup), which look for patterns in the reports. See "Which AI companies we use".
How long: up to 12 months, then deleted. To have one deleted sooner, write to us and tell us what the report said, because we can't link it to you.

"Was this check useful?"

Under a Tip Checker result you can answer Was this check useful? Yes / No. It works with or without an account. We save one row (the "tip_feedback" table): the date in India, your answer and the language of the site. We do not save the post you checked, your name, email, account or IP address with it (we only count how often one connection answers, in memory, with a scrambled address, to stop spam). We show the total on our public track record page. Answers are deleted after 365 days.

The public record of Tip Checker checks

When an AI check finishes, we save one row about it for the public track record (the "tip_check_log" table): the date (India time, without the time of day), the companies the post named (from our company list), which kinds of red flag were found, how much hype the AI saw (high, some or low), and how many news claims matched, contradicted or were not found in our sources. We never save the post or the screenshot itself, and nothing about who ran the check: no account, device code or IP address. If we later find a check was wrong, we mark it as wrong on the public list with a short note. These rows are kept so the record stays complete.

Keeping and deleting your data

We keep your account, saved stories, watchlist, followed IPOs and alert settings until you remove them or ask us to delete them, and waitlist emails until Pro opens or you ask us to remove them. To see, correct or delete your data (including your whole account), email support@hypeless.in from the email address you signed in with. We will do it within 30 days and tell you when it's done. If you delete your account, records of payments made through Cashfree are kept for tax and legal reasons, with your email address removed from them. The anonymous counts described above can't be linked to you, so there is nothing personal in them to delete.

The SEBI registration check

When you look up a registration number or name on our SEBI check page, what you type is sent to our server, which searches SEBI's public register for it. We don't save what you searched or write it to our logs. To avoid asking SEBI the same thing again and again, our server remembers SEBI's public answer for a number or name for up to 24 hours.

The optional Read with AI button (signed-in people only) sends the message you pasted, at most 3,000 characters, through our AI gateway Mesh API to DeepSeek (or Google's Gemini as a backup, see "Which AI companies we use"), only to find the registration numbers and the names in it. The AI says nothing about anyone: the page then looks each number or name up on SEBI's register. We don't send your name, email, account or IP address, we don't save the message, and we only record that an AI call happened (a number of tokens and its cost).

Official filings and official updates

Company pages and the Tip Checker can show a company's official filings from the BSE stock exchange, and some pages show the newest SEBI and RBI press releases. When the Tip Checker looks up filings, your browser sends our server only the company's name from our own list, never the post you pasted. Our server asks BSE, SEBI and RBI for their public lists and keeps their answers for a few hours. No personal data is involved, and nothing about you is sent to BSE, SEBI or RBI.

Information saved on your own device

If you use the site on a phone or add it to your home screen, your browser saves a copy of our pages so they open without internet. This copy stays on your device. You can delete it at any time by clearing this website's data in your browser settings.

When you open the Tip Checker without signing in, our server may also set one small cookie,hypeless_device: a random code (for example "3f9c…") with a signature, so we can count the day's free AI checks per device (5 a day) without an account. Your browser may also keep a similar random code in its own storage. It says nothing about who you are, and we only keep a scrambled form of it in memory until midnight India time. If you are signed in, your AI checks are counted against your account instead. Clearing this website's data deletes it.

Links to other websites

We link to news websites, the Securities and Exchange Board of India (SEBI), the Reserve Bank of India (RBI) and the stock exchanges (BSE, NSE). When you open those links, their own privacy policies apply, not ours.

Children

Our site is designed to be safe for learners of any age, and everything free works without an account. Under India's data protection law, if you are under 18 you need a parent's or guardian's permission before creating an account. That's why the sign-in page asks you to confirm that you are 18 or older, or that your parent/guardian has agreed. We don't track children or show them advertising. If a parent believes their child signed up without permission, email us and we will delete the account. Please also ask a parent before using the AI checks, because they send the text you paste to DeepSeek in China, or Google as a backup (through Mesh API).

If this changes

If we add features that use personal data in a new way (for example, new analytics or the WhatsApp bot), we will update this policy before the feature goes live. We will explain what we collect and why, and ask for your consent where the law requires it.

Questions or complaints

Email us at support@hypeless.in (Grievance Officer: Kriplani, ceo@hypeless.in). We aim to reply within 24 hours and to resolve complaints within 15 days. More details are on our Contact page.

Privacy Policy