Agents do the work.
You approve. The proof is automatic.
AI agents discover your personal data, assess your gaps against the DPDP Act and Rules, propose the fixes — and only after you approve, execute them with a verifiable, rollback-capable, permanently auditable trail.
No live data connectors. No credit card. Real assessment against 46 DPDPA controls in 5 minutes.
Posture
62/100
How it works
Five steps. Agents do the work. You approve at every mutation.
Find personal data across your systems
Categorise by DPDPA type
Score against 46 controls
Typed actions with rollback
You sign. Then execute.
The non-negotiable safety rules
These are enforced architecturally, not by policy. You cannot turn them off.
No mutating agent action executes without a recorded human approval — a signed, scope-bound token, validated per action.
Every executable action needs a completed dry-run with a readable diff before approval is even possible.
Every action carries a generated, validated rollback plan, created at planning time — not under pressure at execution time.
Every action — read or write, agent or human — is written to the append-only, hash-chained audit ledger.
The planning agent (Sudhaar) holds no write credentials. Separation of duties between proposing and executing.
Blast-radius caps are enforced pre-flight and in-flight, with a global kill switch.
Free 5-minute DPDPA gap-scan
Answer 12 questions. Get a prioritised report against the same 46-control library a paid engagement uses. No live data connectors — your inputs stay on this page.
Tell us about your business
Enforcement is 8 months away.
Discovery + remediation take 6–9 months. Starting at enforcement means being non-compliant at enforcement.