What changed in this version (2026-10-14): PlanetScale, Inc. joins the providers in section 3 as the host of our primary database, on Amazon Web Services in the United States; section 4 names that location and section 5 describes its backups. Nothing else changes.
This policy explains what Renaissance Geek, Inc. ("we", "us") collects about the people who use Impeccable PRs (the "Service"), why, and what happens to it. It covers the website at impeccable.reviews, the GitHub App, and the workspace. It does not cover the contents of your repositories and previews as such; those are Customer Content under the Terms of Service, which we process only to provide the Service, as your processor under the Data Processing Addendum.
Plain-language summary: We collect what GitHub tells us when you sign in or install the App, what the App reads to review pull requests, and records of what the Service did. The closed beta is free; billing details would go through Stripe if fees are introduced later. We send review inputs to the AI providers listed in section 3 only under terms that exclude training on them. We use PostHog to understand how the website and workspace are used, including recordings of the workspace. We do not sell data and we run no advertising. Review artifacts expire on a schedule; uninstalling deletes review files, and review records stay until you ask us to delete them.
1. What we collect
Account and sign-in data. When you install the App or sign in, GitHub sends us your login, display name, avatar URL, the installation id, the account you installed on, the repositories selected, and, where the App has the Members permission, whether you own the organization. If you grant email access, we also collect your primary verified email address for account communications, even if it is private on GitHub. We store a signed session cookie in your browser so you stay signed in for up to seven days. We do not store your GitHub access token.
Repository and pull request data. To review a pull request the App reads the changed files, the pull request title and body, repository files needed for context (including design documents and configuration), and the author's login. Pull request authors and commenters appear in review records by their GitHub login. When a repository editor presses a button that asks the App to change the repository (the test pull request during setup, or the proposed update to an outdated preview workflow), the App writes that change to a branch of its own, opens a pull request, and records who pressed the button and what happened.
Preview and persona data. When you connect a preview deployment, tunnel, or uploaded site, we load it in an automated browser and store screenshots, recordings, audio, and text describing what our simulated users did and saw.
Review records. For each review we keep what was posted, when, at what tier, what it cost, the model inputs and outputs needed to produce and display it, and diagnostic logs.
Acceptance records. When an account owner accepts the Terms of Service we record their GitHub login and name, the terms version and its content hash, the time, a salted hash of their network address, and their browser's user-agent string. This is our evidence that the Terms were accepted and by whom.
Billing data. The Closed Beta is free and asks for no payment details. If fees are introduced later, Stripe will collect and store payment details, and we will store the Stripe customer and subscription identifiers, your billing status, and which pull requests were billed. We never see full card numbers.
Waitlist data. If you join the beta waitlist we store your email address and your answers to the waitlist's questions: how your pull requests get previews, whether there is a staging or production address we may use and whether test accounts exist, your front-end stack, how many people open frontend pull requests in a week, your role, whether your team gets its notifications in Slack, and, if you give them, your GitHub organization and what you would want a design review to catch. We also record which priority link you came through, if any, and keep salted hashes of your network address and browser, used only to prevent abuse of the form. Signups from before 2026-10-09 may also hold a short note about what you build. We use the answers to decide whom to invite and in what order. We email you about the waitlist at most twice: once when you sign up, with a link back to the questions, and once when we let you in, with your personal invite. Both have an unsubscribe link.
Usage and analytics data. On impeccable.reviews and in the workspace we use PostHog to record page views, clicks, heatmaps, JavaScript errors, and, once you are signed in, session recordings, along with page and referrer, device and browser details, and approximate location. PostHog stores a random identifier in a cookie and in local storage to link visits from the same browser. When you are signed in, these events carry your GitHub numeric id and login and the account you are viewing. Recordings of the workspace include the page address and can include the text of reviews, persona visits, and buttons, all of which can name your account, repositories, files, and pull requests; they are configured to mask form fields and the other labels that name them, and to leave out code blocks. The Service also sends PostHog records of what it did (installs, repositories added or removed, sign-ins, terms acceptances, and reviews started, posted, skipped, or failed, with the account and repository names and counts involved; and waitlist signups and answers, with only where the signup came from, whether a priority link was used and the resulting group, never the email address or the answers) and reports of errors, which include the error message. These requests go through impeccable.reviews, which forwards your network address to PostHog so it can estimate location. The operator console is not recorded. The permissions and data handling page lists what recordings mask and how long PostHog keeps each kind of data.
Operational data. Cloudflare, which hosts the Service, keeps request logs (network address, user agent, path, timing) for a limited period. Our own service logs record what the Service did for whom, by GitHub login and account, and are kept by Cloudflare for its limited log-retention period. We keep an audit log of actions taken in the workspace and operator console.
2. Why we process it
We process this data to provide the Service you asked for (reviewing pull requests, showing results, running your workspace); to bill you, if fees are ever introduced; to keep the Service secure and detect abuse; to comply with law; to understand how people use the website and workspace so we can improve them (the usage and analytics data in section 1); and to improve the Service using aggregated, de-identified information. Under the GDPR and similar laws, the legal bases are performance of a contract, our legitimate interests in operating, securing, and improving the Service, and compliance with legal obligations. For personal data inside Customer Content, you are the controller and we act as your processor under the Terms.
3. Who receives it
We share personal data only with the providers we need to run the Service, each bound by their own terms and, where applicable, a data processing agreement:
- Cloudflare, Inc. hosts the Service, its storage and workflows, and the automated browser, and sends our email.
- PlanetScale, Inc. hosts our primary database on Amazon Web Services in the United States (Oregon): account and sign-in data, review records, acceptance records, waitlist entries, billing identifiers, encrypted preview credentials (the keys stay with us), and audit logs.
- Google LLC (Gemini API) provides review models, and text-to-speech and image generation for persona visits.
- Anthropic, PBC (Claude API) may provide models that read repositories to configure the Service and to review.
- Cerebras Systems Inc. (Cerebras Inference API) may provide models for selected review steps.
- PostHog, Inc. provides product analytics and session recordings, as described in section 1.
- Stripe, Inc. handles billing, if fees are introduced after the Closed Beta.
- GitHub, Inc. is the platform through which your data reaches us and to which reviews are posted.
We send content to an AI provider only under API terms that state it is not used for training. The providers may retain prompts and outputs for a limited period for abuse monitoring under their own policies. We do not sell personal data, and we do not share it with advertisers or data brokers. We may disclose data if required by law, or to protect the rights, safety, or property of us, our customers, or others.
4. Where it is processed
The Service runs on Cloudflare's global network and our AI providers' endpoints, our primary database is hosted by PlanetScale on Amazon Web Services in the United States (Oregon), and PostHog stores analytics in its United States region; data may be processed in the United States and other countries. Where required, we rely on the providers' standard contractual clauses or equivalent transfer mechanisms.
5. How long we keep it
- Review run artifacts (screenshots, recordings, audio, raw model responses) are configured to expire 90 days after creation.
- Uploaded preview archives are configured to expire 30 days after creation.
- Persona portraits generated for your repositories are kept while the installation is active and deleted by the uninstall sweep.
- Workflow state (the intermediate steps of a review, which can include excerpts of your code) is kept by our hosting provider for its own limited retention period and is not served to anyone.
- Review records (run metadata, findings, scores, and the derived design profiles and personas for your repositories) are kept in our primary database while the installation is active, and after uninstall so that abuse can be investigated and the account restored if you reinstall. We do not delete them on a schedule. Email privacy@renaissance-geek.ai and we delete them within 30 days.
- Terms acceptance records are kept after uninstall, for as long as needed to evidence the agreement.
- When an installation is uninstalled, it is marked deleted, the Service stops reading and serving its data, and we run a deletion sweep over its stored artifacts. Database backups keep deleted rows until they expire: twelve-hourly copies for two days, daily copies for 35 days and monthly copies for twelve months, as the permissions and data handling page lists. Request logs are kept by our hosting provider for a limited period.
- Billing records are kept as long as required for accounting and tax purposes.
- Waitlist entries, with your answers, are kept until you ask us to remove you (email privacy@renaissance-geek.ai) or we close the waitlist. Unsubscribing keeps the entry, marked so that we email it nothing more.
- PostHog keeps analytics data for the periods published on the permissions and data handling page. Email privacy@renaissance-geek.ai and we delete the analytics data linked to you.
These retention periods describe how the Service is configured; details and dates are published on the permissions and data handling page.
6. Review media shown on GitHub
Images we post into your pull requests are displayed by GitHub, which fetches them without a login, so we serve them from unguessable links. The permissions and data handling page lists which images this covers and how the rest of your review media is protected.
7. Cookies
We set a signed session cookie when you sign in to the workspace, a short-lived state cookie during GitHub sign-in, a cookie that grants access to gated preview hosts, and a preference cookie for your theme choice. PostHog, our analytics provider, stores a random identifier in a cookie and in local storage to link visits from the same browser (section 1). We use no advertising cookies.
8. Your rights
Depending on where you live you may have the right to access, correct, delete, or export your personal data, to object to or restrict certain processing, and to complain to a supervisory authority. To exercise these rights, email privacy@renaissance-geek.ai. Uninstalling the App from GitHub triggers deletion of the artifacts of that installation; review records stay until you ask us to delete them, as section 5 says. If your request concerns personal data in Customer Content, we will refer you to the customer that controls it or act on their instruction.
9. Children
The Service is for businesses and professionals. We do not knowingly collect data from anyone under 16.
10. Security
We use administrative, technical, and physical safeguards appropriate for a pre-release service, summarized on the permissions and data handling page. No system is perfectly secure. If we become aware of unauthorized access to personal data we will notify affected customers without undue delay, in the workspace and at the email address on the account, if there is one.
11. Changes
We may update this policy. Material changes are announced in the workspace and at the email address on the account before they take effect. A new provider in section 3 is announced before it receives personal data. The current version is always at https://impeccable.reviews/privacy.
12. Contact
Renaissance Geek, Inc. · privacy@renaissance-geek.ai · security reports to security@renaissance-geek.ai.