Privacy PolicyDraft

What we collect

• Account data: name, email, and organization/workspace membership.
• Your content and derived data: documents you upload and data derived from them to provide the service (extracted text, chunks, embeddings, and generated answers with their citations).
• Usage metadata: plan, entitlement counters (uploads, answers, storage, AI usage), and audit events for security-sensitive actions.
• Support communications: messages you send us.

How we use it

• To provide the service: store and process your documents, generate grounded answers, and enforce per-tenant isolation and access controls.
• To meter and cap usage, maintain security and audit trails, and operate, debug, and improve reliability.
• We do not sell your data, and we do not use your document text, prompts, or answers to train shared/foundation models.

Sub-processors

We rely on a small set of providers to run the service. Each receives only what it needs:

• Hosting, database, object storage, and job queue: Railway (railway.app) - compute, managed PostgreSQL, S3-compatible object storage for uploaded file bytes, and Redis.
• Web hosting / CDN: Vercel (vercel.com) - serves this site and the app frontend.
• Transactional email: Resend (resend.com) - email verification and account messages only.
• AI inference (platform-managed path only): OpenAI (openai.com) - produces embeddings, answers, OCR output, or transcripts when the relevant platform-managed provider path is enabled. Not used on bring-your-own-key or local/self-hosted modes.

A data processing addendum (DPA) is available on request for Early Access participants. This sub-processor list may change as the service evolves; material changes will be reflected here.

AI provider and data handling

• Provider mode is set at the organization level: platform-managed, bring-your-own-key, or local/self-hosted.
• On the platform-managed path, your source text, questions, and uploaded content needed for provider-backed OCR/transcription may be sent to the external AI provider to produce embeddings, answers, OCR output, or transcripts. This is opt-in.
• Bring-your-own-key uses your provider key; local/self-hosted keeps inference on your own hardware. Local/self-hosted is available when the review path requires stricter data control.
• For sensitive or regulated data, prefer bring-your-own-key or local/self-hosted mode and confirm compliance with your institution's policies.

Security

• Each tenant's data is isolated from every other tenant, enforced at the database layer. Access requires authentication and is authorized server-side.
• We do not log your document text, prompts, answers, or secrets in application logs.
• Traffic is encrypted in transit (HTTPS/HSTS). No security control is perfect; report concerns to security@researcher.center.

Retention, deletion, and your rights

• You can access and export your data, and request deletion of documents, workspaces, or your account. Deleting a document removes its derived data (text, chunks, embeddings); account erasure removes or de-identifies your personal data.
• We retain data for as long as your pilot account is active and for a short period afterward for backups and legal/operational needs; backup retention windows are being finalized.
• To exercise these rights during the pilot, contact support. International transfers, legal bases, and retention specifics will be finalized with legal review before general availability.

Children and changes

• The service is intended for researchers and is not directed at children.
• We may update this policy; material changes will be communicated to pilot participants.

Privacy questions? Contact us or email security@researcher.center. See the known limitations for current data-handling scope.