Who runs IEFlow?
The operator and controller of IEFlow product-account data is Top Around Trading LLC, at 1309 Coffeen Avenue, STE 1200, Sheridan, Wyoming 82801, United States. Contact ieflow@victorgurbani.com about privacy and account requests.
What stays out of our membership backend
Your university passwords, session cookies, grades, documents, messages and application contents are not sent to IEFlow’s subscription backend. University connections run locally. Results are passed to the assistant you chose, which may use a remote AI provider. That provider’s practices are separate from ours.
Product-account information
We process your signup email (verified before account access), first name and surname, optional study level/programme, display name, membership status, device labels, trial dates, referral attribution, subscription identifiers and feedback to provide the service, secure accounts and administer billing. Stripe handles payment details; we do not store full card numbers. Supabase provides authentication, database and package storage; Vercel hosts the website.
Purchases use Stripe Managed Payments and Link. Stripe/Link process billing identity, payment, tax and transaction-support information under their own privacy terms. Payment emails come from Link. Your university content and credentials are not supplied to them by IEFlow.
Sign-in and security
Sign-in uses a verified supported email provider or IE address. You can optionally add authenticator protection. Password sign-in is available only with an authenticator and requires both factors. Device access and refresh tokens are stored hashed on our server; your local credential store keeps the usable tokens. Device labels describe your assistant and OS, not hardware fingerprints or personal machine names.
Keeping your connector up to date
Persistent installations check for same-edition software updates while the MCP is running and prepare verified releases for its next restart. You can ask your assistant to disable automatic updates using ieflow_updates. This is software delivery, not optional usage reporting: update checks contain the edition and release channel, not your prompts, university content or tool history. Protected downloads also require membership authentication. Our hosting providers receive ordinary connection metadata, such as IP addresses. Release checks are not counted as installations or downloads.
Basic server traffic counts
We count requests for public website pages to understand whether people are finding IEFlow. This happens on the server, including when optional analytics are declined or JavaScript is disabled. Counters are combined by hour, fixed public page, recognized arrival source (such as WhatsApp), external/internal arrival and browser/known-crawler category. Known link previews are separated from browser opens. Arbitrary campaign values and referrer URLs are not stored: unrecognized sources become “Other”.
These aggregate counters contain no IP address, full user-agent, cookie, browser identifier, account identifier or per-visit record. We do not fingerprint visitors or link page opens into sessions. Reloads can count again; opens are not unique people. Aggregates expire after 90 days. Hosting providers still receive ordinary connection metadata to serve requests. Because the counters cannot identify your requests, they cannot be individually retrieved or erased. The optional controls below govern session and interaction measurement, not these anonymous request totals.
Optional website analytics
Only after you select OK in the cookie notice, we measure sessions, pricing and signup interest, download intent, setup-copy actions and successful signup-email requests. This is separate from MCP usage reporting. A random browser-session identifier lasts up to four hours; reports include fixed public page labels, coarse arrival source and mobile/desktop layout. We do not send URLs, query strings, email addresses, account IDs, chat or university content in these events. Events are not matched to product accounts. They measure observed interest, not exact people or completed installations.
Leave the notice unaccepted to keep optional analytics off. After enabling them, click “Optional analytics: on” at the bottom of the page to withdraw. Switching off stops events and requests deletion of this tab’s current visit; an offline deletion request may not reach the server. We respect Do Not Track and Global Privacy Control. Events expire after 90 days; a short-lived withdrawal marker prevents delayed requests recreating a deleted visit. Hosting providers still process ordinary connection metadata. This choice and its session identifier use local browser storage, only with your choice to enable measurement; no advertising cookies or third-party analytics scripts are added.
Optional usage reporting
Usage reporting is on by default for new accounts. Existing preferences are preserved. Account-free reporting remains off until enabled. Reports contain a random resettable installation identifier, software version, OS family, event or tool name, assistant family, outcome, coarse duration and failure category. We never include prompts, arguments, results, university identities, raw exceptions or filenames. You can disable reporting in your account; doing so deletes the stored raw usage associated with your account. Without an account, you can opt in or out for this installation using the ieflow_usage_reports tool or the ieflow privacy command. Account-free reporting uses a random identifier and a deletion capability kept in your OS credential store. Linking an account ends account-free reporting and removes its old raw reports; your account’s reporting preference takes over. Offline deletion is retried when online. Raw events expire after 30 days. A short-lived revocation record prevents delayed reports from recreating deleted data.
Package download totals are aggregate request counts, not a count of people or installed devices. For linked accounts, successful IE service connections are recorded by service name and date, not university identity or content, to provide trials and understand account setup. Linked-client check-ins support session management; they are not a log of university activity. Infrastructure providers necessarily process network information to deliver and secure requests; this is not a claim that no provider ever sees an IP address.
Cookies and local storage
We use essential authentication cookies and local storage for your sign-in, a referral you chose to follow (for up to 30 days before account linkage) and account setup. No advertising cookies are used. Fonts are served with the website rather than fetched from a third-party font service.
Retention and your choices
Device requests expire after ten minutes and are cleaned up; revoked/expired device sessions and short-lived rate-limit records are periodically removed. Account information is retained while you use the service. Feedback and billing records may need different retention periods; legally required transaction records are not erased with a simple profile deletion.
You can export product-account data, revoke devices and request deletion from your account. We will confirm completion and explain any information that must be retained. Depending on the applicable law, you may also have rights to access, correct, restrict or object to processing, withdraw consent, portability and complain to a supervisory authority.
International providers and updates
The product database is hosted in Ireland. Supabase, Vercel, Resend and Stripe may process service data internationally under their applicable data-processing terms and transfer mechanisms, including standard contractual clauses where required. We use account data to perform our contract with you, security records to protect the service, billing records to administer payments and meet legal obligations, and optional usage reports according to your reporting preference. Material changes will be disclosed here; optional analytics will not be enabled silently.