Data Processing Addendum (DPA)
Template DPA describing how SutraAI processes personal data on behalf of tenants. This is a summary for transparency; a signed DPA should be executed with legal counsel.
Roles
The tenant is the data fiduciary/controller. SutraAI is the processor, acting only on the tenant\u2019s documented instructions.
Nature & purpose of processing
Hosting the tenant\u2019s knowledge base and providing AI-assisted question answering, retrieval, and (where enabled) live-system actions on the tenant\u2019s behalf.
Sub-processors
- AWS Bedrock (model inference) — region ap-south-1 (India).
- Amazon Titan embeddings — for retrieval indexing.
- Managed database, vector store, and cache infrastructure.
Tenants may bring their own model provider, in which case that provider becomes the sub-processor for inference.
Safeguards
Encryption of secrets, RBAC + MFA, per-tenant isolation, PII masking, prompt-injection defenses, audit logging, and response lineage. Primary data is hosted in India (AWS ap-south-1).
Data-subject requests & breach support
SutraAI provides tooling for tenants to fulfil access/export, correction, and erasure requests, and to record and flag suspected data-breach incidents for follow-up.
Contact
Data Protection Office — privacy@sutra.ai
