Fluent privacy and data lifecycle policy

Privacy Policy

Version 2026-05-05.1 · Effective May 5, 2026

This policy covers Fluent early access and the meetfluent.app waitlist. Data you keep in a self-run open-source runtime stays in the environment you operate unless you choose to migrate it into Fluent.

Account data Stored while the account is active, then removed through deletion. Canceled or lapsed accounts are kept up to 90 days for export, support, and reactivation unless deletion is requested earlier.
Runtime exports Export files are handled by the Fluent account or support path and expire after 7 days.
Deletion Waitlist-only deletion starts by email. Provisioned account deletion is operator-assisted today and normally completes within 30 days.

Who Runs Fluent

Fluent early access and the meetfluent.app waitlist are operated by 1001597427 ONTARIO INC., an Ontario corporation using the Meet Fluent brand. For any privacy, data, or policy question, contact us at [email protected].

Geographic Scope

Fluent early access is currently offered to residents of the United States and Canada. We do not actively offer the service to residents of the European Economic Area, the United Kingdom, Switzerland, or other jurisdictions outside the US and Canada. If you submit the waitlist from outside this scope, we may decline your request and delete the submission. Once we expand availability, this policy will be updated and additional notices and rights will apply.

What Fluent Stores

For approved early-access accounts, Fluent stores only the account and domain data needed to operate the service:

What Fluent Does Not Store

AI Assistants And Connected Clients

Fluent is designed to be used through an AI assistant such as Claude, ChatGPT, Cursor, Codex, or Claude Code. When you connect an assistant to Fluent through the Model Context Protocol, the assistant authenticates with OAuth and acts as a client on your behalf. While that session is active, the assistant can read and write the domain data you have authorized.

This means content from your Meals, Style, or Health domains is sent to the assistant provider you chose when you ask the assistant to help with that domain. The assistant provider, not Fluent, controls how that content is processed, retained, and whether it is used for product analytics or model training. Their terms and privacy policy apply to that processing. We recommend reviewing the assistant provider's data and training policy before connecting it.

You can revoke an assistant's access at any time by removing the OAuth consent in Fluent. Revocation stops new reads and writes; it does not retroactively recall content the assistant has already received.

Domain Memories And Derived Data

Domain memories are not a separate hidden profile. They live in the same domain tables that power the product: meal memory, pantry and grocery state, closet profiles and provenance, Health plan state, reviews, and similar records. Some records are derived from earlier user-authored data, such as summaries, source snapshots, or planner state.

When a full account is deleted, derived memories tied to that tenant are deleted with the account data. Fine-grained self-serve erasure of every derived record is not complete yet. If a user needs a specific derived memory removed or recomputed before full account deletion, contact Fluent for operator-assisted review.

Style Photos And Likenesses

Style item photos may incidentally include images of people. Fluent does not run face recognition, biometric identification, or biometric-template extraction on closet photos. Photos are stored as item attachments and are linked to your closet profile and provenance metadata. Do not upload photos of other people without their permission. If a photo of you needs to be removed before full account deletion, contact us and we will action the removal.

Sub-Processors And Service Providers

Fluent relies on a small number of third-party service providers to run the service. The current list, including their role, location, and the privacy policies that apply to their handling of data, is maintained at /subprocessors/. Material additions or changes to this list will be reflected on that page and, where required, called out in this policy.

Cookies And Similar Technologies

The meetfluent.app landing site does not set advertising or cross-site tracking cookies and does not run third-party advertising or analytics tags. Cloudflare may set strictly necessary cookies for security and performance as part of serving the site. Once you sign in to your Fluent account, that product uses session cookies and OAuth tokens that are required for sign-in and to keep your session active; these are described in Fluent account documentation.

When you submit the waitlist, we record a hashed, non-reversible fingerprint of request signals (such as a truncated address and user agent) for rate-limit and abuse protection. We do not store the raw IP address or raw request headers in the waitlist signup row.

Backups And Snapshots

Fluent early access runs on Cloudflare infrastructure. Cloudflare D1 Time Travel keeps database recovery history for up to 30 days on paid Workers plans and 7 days on free Workers plans. Deletion removes data from active Fluent systems first. Backup and recovery copies are not self-serve, are used only for resilience or incident recovery, and age out through the infrastructure retention window rather than by manual user download.

Current export files are runtime artifacts and expire after 7 days. This landing site does not provide a waitlist export endpoint. Open-source runtime snapshot files are created only when an operator runs the snapshot export command in their own environment.

Logs

Fluent writes product audit rows for lifecycle events such as access review, onboarding, export requests, export downloads, and deletion transitions. Runtime logs may include operational errors and routing metadata needed to debug the managed service. They are not a user-facing system of record and are not included in user exports today.

Current product code does not implement a per-user self-serve runtime-log export or deletion path. Product audit rows tied to an active account are removed with account deletion unless a retention exception applies. Runtime-log review, incident records, or security records are operator-assisted.

Waitlist Data

The meetfluent.app waitlist stores:

Waitlist data is retained while Fluent is managing early access, unless you ask us to delete it earlier. If your waitlist record becomes an invite or account, matching access records may also be created in the runtime product.

Personal Information We Collect (CCPA Categories)

For users in California and other US jurisdictions with similar laws, the categories of personal information we collect map to the statutory categories as follows. We have not sold personal information and have not shared personal information for cross-context behavioral advertising in the past twelve months.

Category Examples we collect Purpose Disclosed to
Identifiers Email, account id, tenant id, OAuth subject identifiers Provide and secure the service; sign-in; account communications Hosting, email, auth sub-processors
Customer records Display name, profile, billing references Account and billing Hosting, payment processor
Commercial information Subscription status, checkout session references Billing and access lifecycle Payment processor
Internet or network activity Hashed rate-limit fingerprint, sanitized referrer, request counts Security and abuse protection Hosting
Geolocation (coarse) Country derived at request time for geographic-scope enforcement; not stored long-term Limit signups to in-scope regions Hosting
Inferences Domain memories, summaries, planner state derived from your inputs Provide the Meals, Style, and Health features you authorize Connected AI assistants you authorize
Sensitive personal information Health-domain data (workouts, body metrics, plans); OAuth access tokens Provide the Health domain and authorized connections; never used for inferences about you for advertising Connected AI assistants you authorize

Legal Basis For Processing

For users in jurisdictions that require us to identify a lawful basis (such as Quebec under Law 25), we rely on the following bases:

Your Rights

Depending on where you live, you have one or more of the following rights with respect to the personal information Fluent holds about you:

To exercise any of these rights, email [email protected] from the address associated with your account or waitlist submission, or, if you no longer have access to that address, give us enough information to verify your identity proportionate to the sensitivity of the request. We will acknowledge your request within 10 business days and respond substantively within 45 calendar days. If we need more time, we will tell you why and may take an additional 45 days. You may also have the right to lodge a complaint with your local privacy regulator (in Canada, the Office of the Privacy Commissioner of Canada or your provincial regulator; in the United States, your state attorney general).

Retention Periods

Export Process

Provisioned users can request an export from their Fluent account or support path. The export is a portable JSON file with account metadata, domain memories, Meals data, Style data, Health data, artifact metadata, settings, consent records when available, timestamps, excluded-data notes, binary-artifact handling, and migration notes. This landing site only stores waitlist intake and does not provide a waitlist export endpoint.

If the authenticated runtime export is disabled or paused, the request becomes operator-assisted. A Fluent operator can run the managed export fallback and provide the same logical export categories.

Deletion Process And Timeline

Waitlist-only users can request deletion by emailing Fluent from the landing confirmation flow. Signed-in deletion belongs to the Fluent account path, where users can start account deletion after invite acceptance or onboarding. Provisioned Fluent account deletion moves to operator review until the full multi-table and artifact purge job is self-serve.

Security

Fluent applies technical and organizational measures intended to be appropriate to the risk of the data we hold. Production traffic is served over TLS. Account secrets are stored hashed or encrypted at rest. Access to production data is limited to the operating team and is gated by the authentication and authorization controls of the underlying infrastructure. Database changes are versioned through migrations under source control. Object storage is read-restricted and signed URLs are short-lived. No system can be guaranteed fully secure; if you discover a vulnerability, please report it privately to [email protected].

Breach Notification

If we become aware of a security incident that has compromised the confidentiality, integrity, or availability of personal information we hold about you, we will notify you without undue delay where the incident creates a real risk of harm, and we will notify the appropriate regulator where required by law. The notice will describe what happened, the categories of information involved, the steps we are taking, and what you can do.

Children And Minimum Age

Fluent is not directed to children. You must be at least 16 years old to create an account or join the waitlist. We do not knowingly collect personal information from children. If we learn that we have collected personal information from a person under 16, we will delete it. A parent or guardian who believes a child has submitted information can contact us at [email protected].

Demo And Mock Data

Public Fluent demos use seeded fake data and labeled static mock frames. They are not a user's Fluent account data. Private capture folders, when present, are internal asset-preparation material and should not be published unless they have been redacted and approved under the demo asset process.

Migration

Self-run to early access

The open-source runtime can create local snapshots. Migration into early access is feasible only as an operator-assisted process today, because tenant/profile identity and artifact handling need to be mapped deliberately. Do not assume uploading a snapshot into Fluent is self-serve.

Early access to self-run

The export is portable JSON, not a direct open-source runtime snapshot import. Where feasible, a Fluent operator can use the export fallback and map the exported sections into a migration package. Binary artifacts require separate handling because exports include metadata rather than inline bytes.

Governing Law

This policy is governed by the laws of Ontario and the federal laws of Canada that apply there, except where mandatory consumer-protection or privacy laws of your place of residence require otherwise. Nothing in this policy limits any non-waivable right you have under applicable law.

Changes To This Policy

We may update this policy as the service evolves. The version string and effective date at the top of this document identify the current version. Material changes will be communicated by updating the version, and where appropriate, by emailing the address on your account or waitlist record before the change takes effect. Continued use of the service after the effective date of an update constitutes acceptance of the updated policy. Prior versions are preserved in source control and can be requested by email.

Contact

Export, deletion, migration, rights requests, security disclosures, and any other privacy questions can be sent through Fluent Support or emailed to [email protected].