Trust & security

Security at Marlow

Marlow holds how your organization actually works. This page explains, in plain terms, how that knowledge is protected — what we do, what our platform provider does, and what stays your responsibility.

Last updated August 1, 2026

Overview

This page is maintained by Marlow to answer common security and privacy questions about the Marlow application. It describes the controls that are in place today and is updated as the product changes.

Security is an ongoing process. As Marlow evolves, this page will be updated to reflect new capabilities and practices.

Where Marlow is today. Marlow is currently in public beta and is operated by its founder using managed cloud infrastructure. We deliberately describe what exists rather than an enterprise compliance program: there is no third-party audit, certification, or attestation behind this page, and where it describes platform capabilities those are factual descriptions of features we rely on.

Encryption in transit and at rest

  • In transit: every request to Marlow — the app, the API, and our custom domains — is served over HTTPS with TLS. Certificates are provisioned and renewed automatically by our hosting platform, and HTTP requests are redirected to HTTPS.
  • At rest: application data lives in a managed Postgres database and managed object storage provided by our platform, which encrypts stored data at the storage layer. Backups are stored encrypted.
  • Passwords: passwords are managed by our authentication provider and are never stored or viewable by Marlow.

Hosting and infrastructure

Marlow runs on Lovable Cloud, a managed platform that provides hosting, a managed Postgres database, authentication, file storage, and backend functions. We do not operate our own servers, and no production data is stored on personal devices.

  • The production database is managed, patched, and monitored by the platform provider.
  • Production data is stored in the United States. See Data transfers in our Privacy Policy.
  • Marlow is currently in public beta and is operated by its founder using managed cloud infrastructure. We describe what is in place today rather than a larger program than exists.

Access control and tenant isolation

Every table that holds customer data is protected by row-level security in the database itself, not just in application code. That means a request can only ever return the rows the signed-in user is entitled to — even if a bug existed in the app layer.

  • Access is evaluated at two levels: your workspace membership and role, and the per-map permissions granted on an individual map. A map can be shared with someone who is not a member of the workspace that holds it, and that share grants access to that map alone.
  • Workspace roles determine who can create content in a workspace and who can administer it; map-level permissions determine whether a specific person can view, comment on, or edit that map. Both are enforced in the database, and changes to either are recorded in the workspace and map activity trails.
  • Maps, nodes, comments, questions, and activity are readable only through those two paths — never by simply belonging to the same platform.
  • No customer data is readable by unauthenticated visitors, except content deliberately exposed through a guest share link (see Guest share links below).
  • Sign-in supports email and password plus Google and Apple sign-in. Sessions are short-lived and refreshed with rotating tokens.
  • Access to production systems and customer data is limited and used only when necessary to operate, support, debug, secure, or improve the service.

Backups and recovery

  • The production database is backed up automatically every day, with a rolling retention window of approximately 14 days.
  • Backups are encrypted at rest and restorable by the platform provider.
  • Inside the product, Marlow keeps its own map version history so you can review and restore an earlier version of a map, and it takes an automatic backup snapshot before any restore.
  • Because backups exist, data you delete may persist in encrypted backups for up to about 14 days before rolling out of the retention window.

Vulnerability and dependency scanning

  • Automated security review runs against the codebase and database configuration, including checks for missing row-level security, over-permissive policies, and unsafe database function definitions.
  • Dependencies are scanned automatically for known vulnerabilities, and we upgrade affected packages when findings are reported.
  • Findings are reviewed as part of our normal development work, with critical and high findings taking priority.

Secrets and API keys

  • Credentials and API keys are stored in an encrypted secrets manager provided by our platform, not in source code or in the browser bundle.
  • Secrets are write-only: once saved, their values cannot be read back out of the interface.
  • Secrets are injected only into backend runtimes at execution time and are never exposed to the client.
  • Database service-role credentials are managed by the platform and are not retrievable by us or exposed to the application front end.

Monitoring, logging, and abuse protection

  • Our hosting platform includes automated abuse and traffic-protection measures in front of deployed applications. We do not currently operate a separately configured, dedicated web application firewall; organizations with a specific WAF requirement should contact us.
  • Backend function invocations, errors, and database activity are logged and retained for operational troubleshooting.
  • Automated project monitoring surfaces runtime errors and suspicious behavior to us so we can investigate.
  • Inside the product, workspace and map audit trails record who changed what and when, which gives customers their own record for investigations.

Building trust through product design

Protecting operational knowledge involves both securing data and helping organizations maintain trustworthy documentation over time.

  • Workspace and map permissions: people see and change only what their workspace role and their permissions on an individual map allow, enforced in the database rather than only in the interface, with guest links limited to a single scope and capability.
  • Ownership and accountability: nodes and maps carry an owner, so every documented process has a person attached to it rather than belonging to nobody.
  • Activity history: workspace and map activity records who changed what and when, including permission and status changes, giving you your own audit trail.
  • Version history: maps can be reviewed and restored to an earlier version, and Marlow takes an automatic snapshot before a restore so nothing is lost in the process.
  • Validation and review status: maps and nodes carry status — including a review state with an optional reason — so readers can tell confirmed knowledge from something still being worked out.
  • Questions and assumptions: uncertainty is captured explicitly instead of being flattened into confident-looking documentation.

Subprocessors

We keep our vendor list intentionally small. Current subprocessors that may process customer data on our behalf:

Lovable CloudApplication hosting, managed Postgres database, authentication, file storage, and backend functions.
Transactional email providerDelivery of account, invitation, and notification emails (recipient address and message content only).
StripePayment processing and billing for paid plans. Card details are sent directly to Stripe; Marlow never sees or stores full card numbers.

We will update this list before a new subprocessor begins processing customer data. We do not use advertising or data-broker vendors.

Data retention

Retention is summarized here and described in full in our Privacy Policy.

Active account dataRetained while your account is active.
Account deletionSoft-deleted for 30 days so it can be recovered, then permanently deleted within 60 days of the request.
Archived mapsRetained until you permanently delete them. Permanent delete is immediate and irreversible.
Map version history and auto-backupsRetained according to the current application configuration.
Audit and activity eventsRetained according to the current application configuration.
Database backupsApproximately 14 days, rolling.
Support and email delivery logs12 months.
Billing records7 years, to meet tax and accounting obligations.

Incident response

We investigate all reported security issues as quickly as practical, based on their severity. Critical issues receive our highest priority. If we determine that customer data has been affected, we will notify affected customers and comply with applicable legal notification requirements.

  • Reports and automated alerts come to security@makeitmarlow.com and are reviewed directly by Marlow.
  • We work through containment, eradication, and recovery, and keep an internal written record of what happened.
  • Where a fix or configuration change is needed, it is prioritized ahead of ordinary product work.
  • Where notification is required by law, we notify affected customers and the relevant supervisory authorities.

Report a vulnerability

If you believe you've found a security issue in Marlow, email security@makeitmarlow.com with steps to reproduce. Please give us a reasonable window to investigate and fix the issue before publishing details, and we'll work with you on disclosure timing.

We will not pursue legal action against researchers who report in good faith, avoid privacy violations and service degradation, and do not access, modify, or exfiltrate data belonging to other users. Please do not run automated scanning that degrades service for customers, and do not test against accounts you do not own.

Shared responsibility

Security of the service is a partnership between three parties. Being explicit about which is which avoids the wrong assumption at the wrong moment.

Our platform providerPhysical and network infrastructure, TLS certificates, database and storage encryption, patching of managed services, daily backups, and platform-level abuse protection.
Marlow (us)Application design and code, row-level security policies, role and permission logic, secrets handling, vendor selection, retention practices, and incident response.
You (the customer)Who you invite into your workspaces and at what role, who you grant map-level access to, who you send guest links to and when you switch them off, what information you choose to put into maps, keeping your credentials and devices secure, and removing access when people leave your organization.

Questions about anything on this page? Email security@makeitmarlow.com.