Get started
Security & Privacy

Your competitive intelligence deserves real protection.

Your playbooks, competitive intel, customer stories, and deal strategies are among your most sensitive assets. Here's exactly how Vere keeps them yours.

The most important promise

Stored by Vere. Never used to train AI.

Your intelligence lives in Vere by design — that's how it compounds. It's isolated per organization, encrypted, and yours to delete at any time. When the AI processes a request, it runs on Anthropic's commercial API, which contractually never trains on your data and doesn't retain it long-term. Two layers, one promise: your durable data stays in the Vere platform; the model only ever sees a request in passing.

How the AI processes your data
No model training — commercial API terms prohibit it
Not retained long-term — the provider processes each request in the moment
Zero Data Retention — available for enterprise customers
Access is restricted — governed by Anthropic's commercial terms

This is the AI layer only. Your durable data lives in the Vere platform — isolated and encrypted (see below).

Defense in depth

Security built into every layer.

Database-level isolation

Row-Level Security — PostgreSQL enforces tenant isolation
Defense in depth — blocks cross-tenant access even if the app has bugs
23 protected tables — all sensitive data covered by RLS policies
Audit-friendly — security teams can verify isolation independently

Encryption

TLS 1.3 — all data encrypted in transit
AES-256-GCM — SSO credentials encrypted at rest
AES-256-CBC — OAuth tokens encrypted
bcrypt (12 rounds) — passwords one-way hashed

Authentication

SSO support — SAML 2.0 and OIDC (Okta, Azure AD, Google)
Configurable session length — token expiry adjustable per tenant
Secure cookies — httpOnly, Secure, SameSite=Strict
Account lockout — protection against brute force

Application security

SQL injection protected — parameterized queries throughout
Webhook verification — HMAC-SHA256 with timing-safe comparison
Token revocation — tracked reasons, Redis session invalidation
Server-side entitlements — subscription enforced on every API route

Availability & resilience

Self-healing AI config — auto-recovers from provider changes
Health checks every 6 hours — DB, Redis, job queues, model validity
Zero-downtime deploys — graceful shutdown drains active jobs
Persistent job queues — long-running work survives restarts via Redis

Infrastructure partners

Anthropic commercial API — enterprise-grade AI, no training on your data
PostgreSQL — with Row-Level Security enforcing isolation
Redis — durable job queues and session management
Common questions

The questions security teams ask first.

Will my competitive intelligence be used to train AI models?

No. Never. Vere uses Anthropic's commercial API which explicitly prohibits using customer data for model training. This is a contractual guarantee, not just a setting you can toggle. Your sales playbooks, competitive intel, and deal strategies stay yours alone.

Can other Vere customers see my data?

No. Impossible by design. We implement Row-Level Security at the PostgreSQL database layer. This means even if our application code had a bug, the database itself would refuse to return data from other organizations. Your data is cryptographically isolated.

Questions?

Bring us your security questionnaire.

Book a demo and we'll walk your security team through the details — or hand us the questionnaire and Vere will answer it.

Get started free