August 3, 2026
Subprocessors
Lola engages the vendors below to operate the service, each under a data-processing agreement. This list is distinct from the integrations you connect to feed in your own data.
| Vendor | Function | Data it processes |
|---|---|---|
| Amazon Web Services (us-east-1) | Core infrastructure: storage (S3), key management (KMS), queues, compute, secrets, inbound email (SES), OCR (Textract), transcription (Transcribe), and LLM inference (Bedrock) | Encrypted document bytes, job metadata, encrypted secrets; OCR/transcription/Bedrock process content in-region with no provider retention or training by default |
| Anthropic | Primary LLM (Claude) for extraction, triage, and playbooks | Contract-derived text on the synchronous, zero-data-retention API only |
| OpenAI | Embedding vectors for search | Clause / knowledge / query text under no-train terms |
| Neon | Managed PostgreSQL database | All structured data, extracted text, and encrypted credentials |
| Auth0 (Okta) | Identity provider | Names, work emails, authentication events — no document content |
| Vercel | Frontend hosting | Session cookies transit; API responses render — no document content stored |
| Resend | Outbound transactional email | Recipient addresses, workspace / connector names, and links — never document content |
| Apollo · People Data Labs · TheirStack | Optional company enrichment during onboarding | Public company domain only; never customer content |
Planned / optional adapters not active in production: Google Vertex, Voyage AI, and Azure OpenAI (alternate model routes). We will update this page before adding any new subprocessor that processes customer data.