Run regulated work
on software built for it.
Operating since 2014, Manifestly is SOC 2 Type II compliant, and trusted by thousands of users in financial services, legal, operations, and IT teams across 20+ countries.
Independently audited. Continuously.
Annual third-party audit against the AICPA Trust Services Criteria, plus GDPR and CCPA processing standards.
SOC 2 Type II
Manifestly is audited annually by an independent third party against the AICPA Trust Services Criteria — Security, Availability, Processing Integrity, Confidentiality, and Privacy.
- Last audit
- March 11, 2026
- Auditor
- INTERCERT CPA LLC
- Report
- Available by request in the Trust Center
GDPR & CCPA
Manifestly processes customer data in accordance with the EU General Data Protection Regulation and the California Consumer Privacy Act.
A Data Processing Addendum is available on request to any customer subject to these regulations.
Request DPAEncrypted in transit, encrypted at rest, backed up by default.
Modern cryptography for everything that moves and everything that's stored. Backups, residency, and recovery are documented in the Trust Center.
Encryption in transit
All data between customer devices and Manifestly is encrypted using TLS 1.2 or higher.
Older protocols and weak cipher suites are disabled.
Encryption at rest
Workflow content, attachments, and audit logs are encrypted at rest using AES-256.
No customer data is stored on disk in plaintext.
Backups & recovery
Heroku Postgres (Premium) with point-in-time recovery, plus scheduled encrypted backups to Amazon S3.
Single sign-on, role-based control, and a complete audit trail.
Control who can do what — and keep an exportable record of who did what, when.
SAML Single Sign-On
Available on enterprise plans. SAML 2.0 works with any compliant identity provider.
Role-Based Access Control
Granular permissions by department, role, and workflow. Administrators control who can:
- Author and edit SOPs
- Run workflows and assign tasks
- View completion history and audit logs
- Export data
- Manage users and integrations
Audit logs
Every action logged with the acting user and timestamp. Retained for the life of the account, exportable for compliance review.
- Workflow creation, edits, and version changes
- Workflow runs, step completions, and skips
- Data field changes
- User and permission changes
- Data exports
- Integration and API activity
Security work, every commit and every release.
Code review, static analysis, dependency scanning, and an annual third-party penetration test.
Secure development lifecycle
- Code review required for every production change
- Static analysis (SAST) runs on every build
- Dependency scanning for known vulnerabilities
- Pre-production isolated from production data
Vulnerability management
Continuous monitoring of third-party dependencies. Patch SLAs:
Penetration testing
Annual third-party penetration tests covering the application and supporting infrastructure.
Responsible disclosure
Found something? Report it to security@manifest.ly. We acknowledge reports within 2 business days.
Built on Heroku, on AWS. Inheriting their controls.
Manifestly inherits the physical security, environmental controls, and certifications of Heroku and AWS — SOC 2, ISO 27001, PCI DSS.
Cloud hosting
Manifestly runs on Heroku, which operates on Amazon Web Services.
Network security
- TLS-terminated load balancers
- Internal services on Heroku private network
- Databases not publicly exposed
Data residency
Customer data is hosted in the United States.
Built on our security model. Not bolted onto it.
Manifestly's MCP server lets AI agents — Claude, ChatGPT, Microsoft Copilot, Gemini Enterprise — work with your workflows, SOPs, and assignments. Same permissions. Same audit trail.
AI inherits user permissions
Connected agents can only perform actions the authorizing user is authorized to perform. No elevated privileges, no service accounts with broader access, no shadow paths around RBAC.
AI actions appear in the audit log
Actions performed by AI agents are recorded under the user who authorized the connection and flagged as AI-initiated.
Explicit, revocable connections
Users connect AI agents through an authenticated flow. Connections can be revoked at any time by the user or by an administrator.
Administrative controls
Enterprise administrators can disable AI features entirely for their organization.
Your data. Processed only to provide the service.
Clear ownership, a standard DPA, a maintained subprocessor list, and a published retention schedule.
Data ownership
Customer data belongs to the customer. Manifestly processes it solely to provide the service and as directed by the customer.
Subprocessors
Current list maintained in the Trust Center. Advance notice of additions per DPA terms.
View listDeletion requests
Specific records or full account deletion at any time. Verified requests honored within statutory timeframes.
The boring habits that keep the doors locked.
Tight access, vetted staff, recurring training, and a defined response plan when something goes wrong.
Employee access
Production access to customer data is restricted to Manifestly's co-founders. Access requires MFA.
Background checks
All employees with access to customer data complete background checks prior to employment, where permitted by local law.
Security training
All employees complete training on hire and at least annually. Engineering staff receive additional secure-coding training.
Vendor management
Third parties with access to customer data are reviewed before engagement and reassessed annually.
Incident response
Documented incident response plan. Affected customers are notified within 48 hours of confirmation.
Found a vulnerability? Tell us.
We work with researchers in good faith and credit them on request.
security@manifest.ly- Acknowledging your report within 2 business days
- Investigating in good faith with reasonable status updates
- Not pursuing legal action against researchers acting in good faith
- Crediting researchers (with permission) once issues are resolved
Everything procurement, security, and compliance teams ask for — in one place.
Pull the latest report, monitoring posture, and subprocessor list from our Trust Center — no NDA round trip required.