Trust & Data Boundaries

Evidence first. Controlled side effects.

Trust isn't a feature we bolt on. It's the architecture. Every layer of Verachi is built around strict data isolation, conservative defaults, and complete transparency.

Architecture principles

How we protect your data.

Isolation

Multi-tenant by design

Every workspace is isolated at the database level using Postgres Row-Level Security (RLS). Your data never co-mingles with another workspace's — not in queries, not in caches, not in backups.

  • Row-Level Security enforced on every database query
  • Workspace-scoped API keys and sessions
  • Isolated search indexes per workspace
Encryption

Protected at every layer

Data is encrypted at rest and in transit. Integration credentials are stored with an additional layer of application-level encryption beyond the database.

  • AES-256 encryption at rest
  • TLS 1.3 for all data in transit
  • OAuth tokens encrypted with per-workspace keys
Philosophy

Insight, not surveillance

Verachi captures decisions and context. We have no concept of employee productivity, keystrokes, or individual performance metrics. There are no manager dashboards comparing team members.

  • No productivity tracking of any kind
  • No individual performance scoring
  • No data sold to third parties
Integration permissions

What we access — and what we don't.

Verachi requests the minimum OAuth scopes needed. Here's exactly what each integration accesses.

Integration What we read What we write What we never access
Slack Messages in channels the bot is invited to; thread replies; reactions Messages posted by the Verachi bot (only when triggered) DMs, private channels (unless bot is explicitly invited), file uploads, user presence
Jira Issue titles, descriptions, comments, status changes, project metadata Labels or comments (only via configured rules, reviewable in audit log) Tempo logs, personal boards, sprint velocity, individual workload
GitHub PR titles, descriptions, review comments, issue discussions None by default; optional PR comments via configured rules Source code contents, commit diffs, CI/CD secrets, Actions logs
External actions

Conservative write-backs.

Verachi reads far more than it writes. When we do act on your behalf, every action follows these rules.

Rule 1

Explicit

You initiate the action, or a workspace admin configures the automation rule. Verachi never writes to your tools without a clear trigger.

Rule 2

Reviewable

Every external write is logged in a user-visible audit trail. Any workspace member can see what Verachi did, when, and why.

Rule 3

Reversible

Write-backs are designed to be non-destructive. Labels added, comments posted — never edits to existing content, never deletions.

Transparency

User-visible audit trail

Every action Verachi takes — every data sync, every write-back, every AI-generated summary — is recorded in an audit log accessible to all workspace members. There are no hidden operations. You see exactly what we see.

Compliance

Standards and certifications.

In progress

SOC 2 Type II

We are currently pursuing SOC 2 Type II certification covering security, availability, and confidentiality trust service criteria.

Supported

Data residency

Enterprise customers can choose data residency in the US, EU, or Asia-Pacific regions. Contact sales to discuss your requirements.

Questions about security?

We're happy to walk through our architecture, answer your security questionnaire, or set up a call with our engineering team.