Privacy Policy

Last updated: April 2026

What We Collect

We collect account information (email, name), workspace configuration, API usage metadata (request counts, token counts, latency, model used, cost), and billing information. We do not collect or store prompt content or model responses unless you enable tracing retention.

How We Use Data

Usage metadata is used to provide the dashboard, calculate savings, route requests, operate the cache, and generate invoices. We do not sell your data. We do not use your data to train models.

Data in Transit

Skalpel processes requests in transit to apply caching, routing, and compression. Request content passes through our gateway and is forwarded to the selected model provider. Content is not persisted beyond the request lifecycle unless tracing retention is enabled by you.

Cache Data

Cached responses are scoped to your workspace, project, and environment. Cache entries are subject to configurable TTL and can be invalidated on demand. Semantic cache candidates are stored as embeddings only.

Data Retention

Usage metadata is retained for the duration of your account plus 90 days. Tracing data follows your configured retention policy. Billing records are retained per legal requirements. You may request deletion of your account and associated data at any time.

Enterprise

Enterprise customers have access to configurable data retention, optional no-body logging, PII redaction in traces, and self-hosted deployment options for complete data control.

Contact

For privacy inquiries, contact privacy@skalpel.ai.