Security

Security built around
workspace isolation.

ONMO verifies identity, workspace membership, and authorization on every authenticated API request. Customer data, advertising credentials, and assets remain scoped to the workspace that owns them.

  1. 01Authorization on every request
  2. 02Workspace-scoped data access
  3. 03Credentials encrypted with AWS KMS

Authorization

Access is verified
on every request.

Authorization is re-verified on every authenticated API request.

Removing access should mean removing access.

ONMO re-verifies account status and workspace membership on every authenticated API request rather than relying on a cached authorization decision. If a teammate is removed from your workspace or their account is disabled, subsequent requests no longer inherit their previous access.

Identity decides, not the request.

Authorization is determined from the user’s verified identity and workspace membership, not from a workspace identifier supplied by the request.

Workspace isolation

Every workspace
is isolated.

Every authenticated request is scoped to the workspace contained in the user’s verified identity.

No workspace switching by parameter.

Requests cannot select another workspace simply by changing an ID or request parameter. If authorization fails, the denied request carries no workspace data forward into the application flow.

Logical isolation throughout the platform.

Customer data is logically isolated throughout the platform, and access controls are designed to prevent one customer from accessing another customer’s data.

Advertising credentials

Encrypted and
workspace-bound.

AWS KMSConnection credentials for Meta, Google, TikTok and Snapchat are encrypted using a dedicated AWS Key Management Service key.

Credentials are encrypted and cryptographically bound to your workspace.

Encrypted credential fields are cryptographically bound to the workspace, advertising platform, and credential field they belong to. A credential encrypted in one context cannot simply be moved to another workspace or credential field and decrypted there.

  • Browser-bound authorization flows with single-use state validation, where the platform flow supports it
  • PKCE where the platform supports it
  • Disconnecting a platform removes the stored connection credentials
  • ONMO never asks customers for their advertising-platform passwords

Encryption

Encryption throughout
the platform.

Data is protected both in transit and at rest.

  • TLS 1.2 or higherRequired for every external connection.
  • HSTSBrowsers are instructed to connect only over HTTPS.
  • Verified TLS to the databaseApplication connections to database infrastructure verify the server certificate.
  • Database encrypted at restDatabase storage is encrypted at rest.
  • Object storage encrypted at restUploaded assets are encrypted at rest.
  • Private buckets, signed accessCustomer assets are stored in private buckets and served through signed, workspace-scoped access.

Infrastructure and identity

Built on AWS.
Identity managed with Amazon Cognito.

ONMO is hosted on Amazon Web Services in AWS us-east-1. Authentication is managed using Amazon Cognito.

Hosting
Amazon Web Services us-east-1
Identity provider
Amazon Cognito Advanced threat protection
Two-factor authentication
TOTP two-factor authentication available to every user
Sessions
Short-lived authenticated sessions
Identity resources
Infrastructure safeguards protect identity resources from accidental deletion

Application layer

Hardened at the
application layer.

Controls a technical reviewer will look for, in place across the product.

  • Content Security PolicyBlocks inline and eval-based scripts.
  • Restricted CORS allowlistsOnly approved origins can call the API from a browser.
  • HMAC-verified webhooksInbound webhooks must carry a valid signature.
  • Constant-time signature comparisonSignature checks do not leak timing information.
  • Rate limiting on AI endpointsGeneration and analysis endpoints are throttled per caller.
  • Outbound request protectionsServer-side fetches of customer-supplied URLs are blocked from internal-network addresses.
  • Workspace-scoped authorizationEnforced on every authenticated request, not per feature.
  • Private storage, signed accessNo public buckets; assets are served through signed access.

Engineering process

Security is part of the
development process.

Security controls are reviewed continuously as ONMO changes.

Every commit and CI run is subject to automated secret scanning. Production builds install dependencies from a committed lockfile, and deployment pipelines do not depend on long-lived cloud access keys.

Dependency security updates are documented and reviewed. ONMO also conducts independent senior engineering security reviews, with findings documented and tracked through remediation.

  • Automated secret scanning on every commit
  • Secret scanning repeated in CI
  • Production builds install from a committed lockfile
  • No long-lived cloud access keys in the deployment pipeline
  • Documented dependency security reviews
  • Independent senior engineering security reviews
  • Findings documented and tracked through remediation

Data control

Your data remains
under your control.

Deletion

Account deletion is available in-product from Profile Settings. Customer data deletion requests, including for connected platform data, can be sent to privacy@onmo.ai. Our connected-platform data policy explains what happens to data when you disconnect a platform.

Backups and recovery

Database backups retain up to 35 days of history with point-in-time recovery.

Your team sets the boundaries

ONMO acts within the goals, budgets, campaign scope and permissions you approve. Changes to goals, scope or limits require your approval.

AI service providers

Where ONMO uses third-party AI services to provide creative-generation functionality, processors may include Anthropic, Google, OpenAI, Adobe, and Runway. Use of these services depends on the feature being used and remains subject to ONMO’s contractual, privacy, and data-handling requirements. See the Privacy Policy and the Data Processing Addendum.

Security reviews and responsible disclosure

One address for
security questions.

security@onmo.ai

Enterprise security questionnaire, architecture, data-handling, or vulnerability disclosure questions can be sent to security@onmo.ai. Responsible vulnerability reports are reviewed and routed to the appropriate engineering owner for investigation.

We acknowledge security inquiries within two business days.

  • Enterprise security questionnaires
  • Architecture questions
  • Data-handling questions
  • Vulnerability disclosures
  • Procurement and security review requests

Have a security question?

If your team needs additional information about ONMO’s architecture, integrations, access controls, processing practices, or enterprise security requirements, contact security@onmo.ai.

Contact Security