Privacy policy
What we collect, why we collect it, who else sees it, and how long we keep it, in plain English, because a policy nobody can read protects nobody.
Last updated 12 August 2026
01Who we are
Khan Terminal Ltd (company number 17390417), registered in England and Wales at 39 Sheffield Road, Slough SL1 3EF, United Kingdom, is the controller of your personal data. That means we decide what is collected and why, and we're accountable for it.
For anything about your data, whether a question, a request or a complaint, email hello@khanterminal.com. A person reads it.
02What we collect
- Your account. Email address and a password. The password is salted and hashed with scrypt. We never store it in a readable form and cannot recover it.
- What you create. Watchlists, portfolio holdings, saved research, notes, documents you upload, and your conversations with the AI.
- Billing. If you subscribe, your email and subscription status. Card details go straight to Stripe and never reach our servers.
- Contact and newsletter. If you write to us: your name, email and message. If you subscribe: your email address.
- Technical data. Your IP address and standard server logs, used to keep the service up and to rate-limit abuse of public forms.
If you agree to it, we run Vercel Web Analytics, which counts page views and referrers so we can see which parts of the terminal get used. We ask before it loads, and if you decline or haven't answered, the script is never requested. It sets no cookies and doesn't follow you across other sites. We run no advertising and no third-party tracking beyond that, and there is no profile of you being assembled here beyond what you deliberately put in.
03Why we’re allowed to use it
| What we do | Lawful basis |
|---|---|
| Run your account and provide the terminal | Performance of our contract with you |
| Take payment for a subscription | Performance of our contract with you |
| Send the newsletter and marketing you asked for | Your consent, withdrawable at any time |
| Count page views to see what gets used | Your consent, withdraw it any time |
| Reply to your enquiry | Our legitimate interest in answering you |
| Keep the service secure and stop abuse | Our legitimate interest in a working, safe service |
| Keep billing records | Our legal obligation under UK tax law |
04AI, and where your documents go
We use several providers and fall back between them for availability. Today those are Groq, Google (Gemini), Cloudflare and Mistral AI. Each processes your content under its own terms in order to return a result.
We do not sell your data, and we do not train any model of our own on your content. We can't make promises on our providers' behalf about their own training practices, which is precisely why we're naming them here rather than hiding them behind “trusted partners”.
A document you upload is also stored, not just passed through: we split it into passages and keep them in a search index (Qdrant) so the assistant can find the right part later. Each passage is tagged with your account, and a search can only ever return your own. Deleting the document, or your account, removes them from the index as well.
Some analysis runs entirely on our own infrastructure and involves no third party at all. The valuation, screening, ESG, Sharia and stress engines are ordinary code, not model calls.
05Who else processes your data
These are our processors. They act on our instructions, for the purposes below and nothing else. We will keep this list current as the product changes.
| Processor | What for | What reaches them |
|---|---|---|
| Vercel | Hosting, content delivery and website analytics | Requests to the site, including IP address and server logs, plus cookieless page-view and referrer analytics |
| Turso | Database | Your account, watchlist, portfolio, saved research, documents and conversations |
| Stripe | Payments and subscriptions | Your email and billing details. Card details go to Stripe directly and never touch our servers |
| Cloudflare (Workers AI) | AI model provider, text embeddings, search reranking, OCR and image reading | Your questions and the document, filing or web passages needed to answer them, plus text and images from documents you upload, for indexing, ranking and search |
| Groq | AI model provider and audio transcription | Your questions and the document or filing passages needed to answer them, plus any audio you ask us to transcribe |
| Mistral AI | AI model provider and document OCR | Your questions and the document or filing passages needed to answer them, plus scanned documents and images you upload, for text extraction |
| Google (Gemini) | AI model provider and embeddings fallback | Your questions and the document or filing passages needed to answer them |
| Google (favicon service) | Company logos on generated presentations | The website domain of the company a presentation is generated for. Requested by our server, not your browser, and only when the logo option is on. No account or personal data is sent |
| Qdrant | Search index for your documents | The text of documents you upload, split into passages, with the numeric representation used to search them. Each passage is tagged with your account and conversation so a search can only return your own |
| Upstash | Shared cache and rate limiting | Answers the assistant has already produced, so an identical question is not re-run, and request counters keyed by IP address |
| Resend | Sending email | Your email address and the contents of the message: sign-in codes, verification links, security alerts and anything you asked us to send you |
Some of these operate outside the UK. Where data is transferred abroad, we rely on the UK's adequacy regulations, or on standard contractual clauses and the UK Addendum as provided in those providers' terms.
Separately, the terminal fetches public data from sources such as SEC EDGAR, Companies House and market-data feeds. Those requests carry a company or ticker, never anything about you.
06How long we keep it
| Data | How long | Why |
|---|---|---|
| Your account and everything in it: watchlist, portfolio, saved research, documents, AI conversations | Until you delete it | It's your work. Deleting your account removes it and everything attached to it. |
| Sign-in sessions | 30 days, or until you sign out | So you aren't asked to sign in on every visit. |
| Newsletter subscription | Until you unsubscribe | It's based on your consent, so it lasts exactly as long as that consent does. |
| Messages sent through the contact form | 24 months | Long enough to handle a question or dispute properly, and no longer. |
| Email verification, 2FA and password-reset codes | Until used or expired | They're single-use and short-lived by design. |
| Billing and payment records | 6 years | UK tax law requires it. This one isn't our choice. |
07Your rights
Under UK data protection law you can ask us to:
- Give you a copy of the personal data we hold about you.
- Correct it if it's wrong.
- Delete it. You can delete your account yourself, which removes it and everything attached to it.
- Restrict or object to how we use it.
- Give you your data in a portable form.
- Stop sending you marketing. Instantly, via the unsubscribe link or your settings.
Email hello@khanterminal.com and we'll respond within one month. If you think we've got it wrong, you can complain to the Information Commissioner's Office at ico.org.uk. We'd rather you told us first so we can fix it.
08How we protect it
- Passwords are salted and hashed with scrypt. We never see or store the original.
- Sessions are opaque tokens in an HTTP-only cookie, not readable claims, and not reachable from JavaScript.
- Everything is encrypted in transit with TLS.
- You can turn on two-factor authentication, and we'd encourage it.
- Your research, documents and portfolio are scoped to your account.
No system is perfectly secure, and anyone who tells you otherwise is selling something. If you find a vulnerability, please tell us at hello@khanterminal.com.
09Cookies
We set exactly one cookie, and it exists to keep you signed in. Our analytics are cookieless and only run if you agree to them. You can change that answer at any time on the cookie policy. We run no advertising and no cross-site tracking. The cookie policy explains it in full, and it's short.
10Children
Khan Terminal is not intended for anyone under 18, and we don't knowingly collect data from children. If you believe a child has given us personal data, tell us and we'll delete it.
11Changes to this policy
If we change how we handle your data, whether a new processor or a new purpose, we'll update this page and change the date at the top. For anything significant, we'll tell you directly rather than hope you re-read the page.