Security

Every claim on this page has something behind it.

Your customers' questions and your own documents stay inside your account, on Indian infrastructure, under India's own data-protection law.

Claim Backing
Data stored in India Database, file storage, OCR and embedding generation all run in AWS ap-south-1 (Mumbai). One exception: generating a reply uses a cross-region inference call, so the question text and matched excerpts may be processed outside India — transiently, never retained or used for training. Full wording in the privacy policy.
Per-account isolation Every document, chat session, guardrail and usage record is scoped to your account. No shared or cross-account visibility.
Answers grounded in your content Replies are generated from passages retrieved out of your own library — not from open-ended model knowledge.
Wrong answers are the exception we design for Our terms don't warrant perfect accuracy, because no AI system can — so every reply cites its source passage, guardrails block risky topics, and anything uncertain becomes a refusal and a callback instead of a confident guess.
Consent-aware by design Configurable notice and consent checkbox on lead capture, with a permanent per-lead record of exactly what was agreed to — built against India's DPDP Act.
Enterprise LLM infrastructure Powered by AWS Bedrock and Anthropic Claude, not a hobbyist API wrapper.
Scoped, revocable API access Server keys are hashed at rest, can be restricted to your IP addresses, and are individually rotatable with immediate effect.

What we don't claim

No certification badges we haven't earned. If your own compliance team needs something specific — a retention change, a signed agreement, a deployment question — ask and we'll answer plainly.

Read the privacy policy Ask a security question