Trust
Security and compliance
Datafor is built so that every dashboard, embedded report, API call, and AI answer respects the same permissions.
Access control
- Access control lists on content; Creator, Reader, and AI user types
- Row policies with a condition builder and per-user system variables
- Table and column hiding
- "Test access" to simulate what a user or role can see
Authentication
- Single sign-on with OAuth2, SAML 2, CAS, and LDAP
- JWT sign-in for embedded users, signed by your backend
- Personal tokens for MCP clients: per user, shown once, expiring, revocable
AI safeguards
- The language model never writes SQL or MDX; Datafor validates and runs every query
- AI answers run under the asking user's permissions, including row-level security
- MCP tools are read-only and expose no raw queries or other users’ data
- You choose the LLM provider; API keys are restricted to admins and never exported
- Per-user question quotas and usage dashboards
Deployment and data residency
- Self-hosted editions run entirely in your environment; your data stays in your databases
- Datafor queries sources live under each user’s permissions
- Audit log of administrative and content actions
Roadmap
- • Datafor Cloud: hosting region, sub-processor list, and backup policy will be published before general availability
- • Independent security assessment and certification plan
- • Public status page for Datafor Cloud
Privacy policyHow we handle personal data.Subscription agreementTerms for Datafor subscriptions.Service level agreementAvailability and support commitments.
Security questionnaire or vulnerability report? Email marketing@datafor.com.cn.
See Datafor on your own data
Book a guided session for your scenario, or start a 30-day trial on your own infrastructure.