Pipeboard Security One‑Pager
Last updated: November 17, 2026
Overview
Pipeboard is designed with security and privacy as first principles. We limit what we collect, safeguard what we store, and give you control over your data.
How we protect your data
- Minimal data collection: We store only what's needed to operate the product (for example, authentication tokens and basic account metadata).
- Encryption in transit & at rest: All traffic uses HTTPS/TLS. Data stored in our database is encrypted at rest using AES‑256.
- Access controls: We apply least‑privilege access and database Row Level Security so users can access only their own records.
- Short‑lived caching only: Advertising data may be cached briefly to improve reliability and performance. We do not warehouse ad data. To speed up Insights, we may store your ad account list and recent 7‑day aggregate metrics with a creation timestamp; this cache is per‑user and protected by RLS.
- Backups: Our database is backed up automatically on a regular schedule. Backups are encrypted and stored securely.
- Your control: You can disconnect or delete your data at any time. Removing Pipeboard from Facebook Business Integrations or deleting your account removes our access.
Authentication & access model
- Sign‑in options: Login via OAuth providers such as Facebook, GitHub, or your corporate SSO (SAML/OIDC).
- Facebook OAuth for Ads access: We obtain the tokens required to read your Ads/Insights and perform permitted actions. Tokens are stored securely and can be revoked by you at any time via Facebook Business Integrations.
- The MCP gateway authenticates with your Pipeboard session/API token. Your Meta token is resolved and injected server‑side only when needed.
- We do not share your Meta credentials with MCP clients or LLM providers.
Third‑party providers
- Payments: Stripe processes payments; we do not store card numbers.
- LLM providers: Your MCP client (for example, Claude or Cursor) sends data to its own model under its settings and policies. Pipeboard itself calls hosted LLM providers (Anthropic, OpenAI) only when you use a Pipeboard feature that needs one, sends only the data that request needs, and does not let those providers train on your data.
- Sub‑processors: The full list is at the end of this page.
AI and model training
- Current stance: We do not use customer content or connected Meta Ads data to train AI models today.
- Future features: If we introduce model‑improving features, we will provide advance notice and an account‑level opt‑out before using customer content for that purpose. Unless you opt out, your content may be used to improve those features. You can opt out at any time. Enterprise agreements can instead require written opt‑in consent before any such use.
Compliance posture
- SOC 2: SOC 2 Type I audit engaged with Thoropass, an independent CPA firm; the report has not been issued yet. A Type II observation period follows the Type I. Our hosting and database providers hold SOC 2 reports today.
- Google CASA: Tier AL1 verified by an App Defense Alliance–authorized lab.
- Penetration testing: Independent test every year. The 2026 test and its retest are complete; the executive report is available on request at trust.pipeboard.co.
- Vendor reviews: We can support your review with this summary, our DPA and your security questionnaire.
Incident response
We investigate reports promptly and notify affected customers consistent with legal requirements and our commitments. Please report issues to the address below.
Sub‑processors
| Sub-processor | Purpose | Location |
|---|---|---|
| Vercel Inc. | Application hosting and global CDN | United States (global edge) |
| Supabase Inc. (on Amazon Web Services) | Managed PostgreSQL database, authentication, file storage | United States (us-east-1) |
| DigitalOcean LLC | MCP gateway hosting | United States |
| Stripe, Inc. | Payment processing, billing, tax handling | United States |
| Anthropic, PBC | Hosted LLM inference (used only on your instruction; no training on your data) | United States |
| OpenAI, L.L.C. | Hosted LLM inference (used only on your instruction; no training on your data) | United States |
| GitHub, Inc. | Source-code hosting and CI/CD | United States |
| SigNoz Inc. | Application observability and log/trace storage | United States |
| PostHog Inc. | Product analytics and feature flags | United States / European Union |
| Customer.io, Inc. | Transactional and lifecycle email delivery | United States |
| Crisp IM SAS | Customer support messaging | European Union (France) |
Contact
ARTELL SOLUÇÕES TECNOLÓGICAS LTDA
privacy@pipeboard.co
https://pipeboard.co
Full details: pipeboard.co/security