Chiarelli Academy Privacy Policy

Last updated: October 20, 2026

Updated to describe the data processed for the Chiarelli AI Setup subscription (billing identifiers, subscription status and period, updater tokens, download records and the package identification mark) and the retention period. The October 19, 2026 version covered the team's reading of feedback with the help of artificial intelligence assistants, always with human review.

1. Who we are

Chiarelli Academy is operated by Chiarelli Labs, registered under Brazilian CNPJ 62.822.123/0001-65. Questions and requests about personal data can be sent to leonardo@chiarelli.dev.

2. Data processed and purposes

To create and protect access, we process your email address, technical authentication and session records, and, when you sign in with Google, the account's stable identifier (sub). Sign-in requests the openid, email, and profile scopes. Google may provide basic profile data in this flow; to authenticate and link the account, the Academy retains the verified email address and identifier. We do not request access to Gmail, Google Drive, or the content of those services.

Access by a code sent via email processes the address provided and the records needed to issue and validate codes and limit attempts. Challenge and session tokens are stored as hashes. Functional cookies maintain the session and temporary sign-in transaction.

To manage purchases and release content, we record the email address, checkout session and payment identifiers, amount, currency, status, and technical purchase events. Payment is processed by Stripe; card data remains with the provider and is not stored by the Academy.

To provide support, prevent fraud, and defend ourselves in potential disputes, we record each course file you request for download, including the date and time, linked to your account. This record is based on the regular exercise of rights and legitimate interest. Only the Academy team accesses this record, which is stored by the vendors listed in section 3.

While you use the course, we may record lessons opened or completed, video position and duration, activity date, and last access. If you submit feedback, we record the rating, message, suggested next topic, email address, and submission date. We also maintain aggregate counts of page views, checkout starts, and downloads.

The Academy team reads and organizes feedback with the help of artificial intelligence assistants, which classify the subject and prepare draft replies. No reply, refund, or decision about you is made by the tool alone: a person reviews and approves it before it is sent. Feedback does not appear in metrics, except in aggregate form.

You may voluntarily provide a preferred name or nickname, which is optional and linked to your account, solely to personalize the private greeting in the member area as part of the requested service. It can be changed or removed in that area by saving an empty value. It is not imported from Google, email, or testimonials, is not sent to analytics or published as a testimonial, and is retained while the preference exists in the account.

In the Course 02 practical journey, we record your account and course identifiers, content mode and version, mission, phase, choices, feedback already displayed, information requests, self-assessments, consequences, recoveries, and isolated practice sessions. We also store the date and time and technical records that identify each request, to prevent duplicates and allow you to resume the journey correctly.

For the purchase journey, we collect only consented first-party events, without name or email in the event. A pseudonymous identifier may be optionally associated with an authorized order. DNT, GPC, objection, or withdrawal stops new collection. Events and links are retained for up to 13 months; we do not reconstruct old journeys or identify you across tabs or devices.

We use this data to authenticate users, verify purchased access, deliver the course, preserve progress, respond to feedback, protect the service against abuse, and understand how the product works.

For the Chiarelli AI Setup subscription, we process the Stripe customer and subscription identifiers and the subscription status, the renewal date, the update token hash, the random package identifier and the download log. The purpose is to deliver and update the paid kit and deter redistribution. The legal basis is performance of the contract and, for the package identifier, the legitimate interest in protecting the kit. Stripe processes payment data; the Academy does not store card data.

When you arrive through a campaign link (with the parameters utm_source, utm_medium, utm_campaign, utm_term, or utm_content), we store a first-party cookie called academy_utm in your browser for up to 30 days with the campaign source. There are two processing activities, both based on legitimate interest (Article 7, IX, of the LGPD), assessed in an internally documented balancing test. The first is aggregate measurement by source, to understand which promotions bring visits and registrations. The second occurs only if you sign up for a launch notification or create student access: the source recorded in the cookie is stored with that registration. If a source is already recorded, it is not replaced. The cookie is not shared with third parties, except for the transfer to Stripe at checkout described below, and is not created when your browser sends the Do Not Track (DNT) or Global Privacy Control (GPC) signal.

Voting and launch notification

The topic selected for the next course is recorded without identification. Email is optional: if you provide it and check the consent box, we store the email address together with the selected topic, consent date, confirmation date, version of the consent text, and, if applicable, the campaign source through which you arrived. We use this email address exclusively to send a launch notification. The notification is active only after confirmation through the link sent, which expires within 48 hours. You can request removal by emailing leonardo@chiarelli.dev.

Announcements about purchased content

If you purchased a course and have active access, we may send emails about updates to the content purchased: new materials, corrections, and what changed and why. We do not send promotions for other products. The legal basis is performance of the contract (Article 7, V, of the LGPD). We process your normalized email address, a keyed hash of that email address, the status of each send, and the message identifier in Resend. Purchases with a refund or an open or lost dispute do not receive announcements.

Every announcement includes an unsubscribe link, and compatible email providers also offer one-click unsubscribe (RFC 8058). Opening the link does not unsubscribe you: you confirm your choice on the page. Access codes will continue to arrive. We retain a record of what was sent for as long as the purchase exists. The unsubscribe is stored only as a keyed hash of the email address, not as the email address itself, and is retained without a time limit so that your choice continues to be respected.

Testimonials

Feedback submitted in the member area is private by default. We publish the comment only with your explicit authorization, without editing and without your email address. The displayed name and role are the ones you choose. You can revoke authorization by emailing leonardo@chiarelli.dev, and removal from the page is immediate.

Artificial intelligence orientation

When orientation is active, Anthropic receives the question, the history kept only in the tab, and excerpts from the catalogue that your access allows you to consult. We do not store the raw conversation, including question and answer, in the database, analytics, or URLs. To operate the service, we retain individual pseudonymous technical records of admission, status, cost, versions, coverage, and authorized sources. These records do not contain the conversation text or student or demand identifiers.

Resolved technical records may be deleted after 90 days once the cost cycle has also ended. Costs that remain unknown and pending operations stay until reconciliation. The anonymous orientation session lasts up to 30 days. A confirmed demand keeps the summary you reviewed, after a basic contact filter, and its associated sources for up to 180 days after the latest editorial review. The receipt displayed at confirmation lets you erase that demand without providing an email address; it cannot be recovered if lost. An unfrozen, unsent announcement draft may be deleted 90 days after its last edit. Frozen or sent revisions are retained as evidence of the reviewed and sent content.

The Academy limits orientation attempts before a query: 5 per hour for anonymous use or a free account and 20 per hour for a buyer or administrator, shared among AI operations. Demand confirmation uses its own limit. Limits use fixed windows and do not prove one unique person across accounts, browsers, or networks. Separate daily and cost limits also apply.

Anthropic processes requests through its API under the applicable terms and policies and the organization and model conditions in use when the query is made. Retention may vary under those conditions. The Academy sends only what is needed for the response and authorized references. Do not send clients' personal data, credentials, or confidential information.

Formation notebook

The notebook stores the answers and notes you save in the formation. Only you can access that content. It stays available to read, export, and delete even after access to the formation ends.

A Nucleo review includes the saved section and only the notes you check. Its purpose is to guide what you wrote. A failed review does not block saving, exporting, or deleting the notebook.

Deleting the notebook removes the answers, the notes, and the review text. Technical receipts follow that lifecycle and disappear with it. No number of days is promised for this deletion, because there is no separate executor. Closing the account still uses the existing support path.

3. Vendors

We use Google for sign-in, Resend to send codes and email announcements, Stripe for payments, Neon for the database, and Vercel for hosting and access metrics. Each vendor may process the data needed for the function it performs.

We also use Grok Bot as the team's artificial intelligence assistant platform, to read, triage, and draft replies to feedback, and Anthropic when text processing goes through its services. These vendors may process data abroad, including in the United States.

At checkout, if a campaign source is stored, it is passed to Stripe as UTM parameters on the payment link, together with the email address and language already used to prefill the purchase.

Vercel Web Analytics operates only in production and receives events only from authorized public pages. After filtering, the URL may retain utm_source, utm_medium, and utm_campaign. Other parameters and fragments are removed. API routes, the member area, administration, and materials do not generate these events.

Google Analytics and Vercel Web Analytics are optional: they start only after your explicit acceptance on public pages in the Academy's production version. They measure page views, interactions, and campaigns for aggregated reports. The choice remains in the browser's local storage for 180 days. Google uses first-party cookies, without cross-domain sharing, for 180 days without renewal by the Academy. Vercel collects these metrics without cookies.

We do not enable Google Signals, advertising, user_id, or granular location and device collection. GA4 is configured to retain user and event data for 2 months, without resetting on new activity. See Google's Privacy Policy. Vercel's retention periods and criteria follow its privacy policy. These vendors may process data and make international transfers under their own applicable terms.

The Meta Pixel is optional and starts only after your acceptance. It runs only on the home page, courses, about, how it works, privacy, terms, and /comprar. It measures public visits for advertising and ad audiences and may use _fbp and _fbc cookies, plus IP address, browser data, and filtered utm_source, utm_medium, and utm_campaign parameters. We do not send purchase data, email, name, or student data and do not enable advanced matching. There are no events in members, administration, API, materials, or code-bearing pages. Retention and transfers follow the Meta Privacy Policy.

4. Retention and your rights

DNT and Global Privacy Control (GPC), browser privacy signals, take priority: when active, optional metrics are not sent and the campaign source is not captured. You can withdraw your acceptance under "Analytics preferences" in the footer. Withdrawal stops future collection but does not delete data that providers may already have received under their own policies.

The campaign source cookie lasts for up to 30 days. A source recorded with a launch notification or student registration remains for as long as the registration exists. To object, use a browser with DNT or GPC enabled, delete the academy_utm cookie, or request deletion through the email address below.

To prevent your campaign source from remaining in your registration, email leonardo@chiarelli.dev with the subject "Não quero que minha origem seja registrada". We delete the recorded source within 15 days and confirm by email. A source already passed to Stripe in a checkout is governed by Stripe's policy. After the objection request is fulfilled, the source will no longer be passed to Stripe in checkouts made with the email address used for the request.

Data, including journey records, is retained for as long as needed to operate access, record the purchase, preserve security, and meet applicable requirements. The download record is retained for as long as the purchase record exists. We do not state a single retention period or automatic deletion here. You may request confirmation of processing, access, correction, information about sharing, and deletion where applicable by emailing leonardo@chiarelli.dev. The identity of the person making the request may be verified before the request is handled. Feedback content read by the artificial intelligence assistants is retained by the vendors under each vendor's terms. Our feedback tracking records do not store the email address, name, or submitted text.

Subscription data is retained while the subscription is active and for the period already used for other purchase records, as described in the previous retention paragraph.