Pipeline output is written to storage you own (S3 / Azure / GCS). Your source data is never sold and never used to train models.
Sensitive fields are detected automatically and masked in the Silver layer — salted SHA-256 or nulled for high-risk identifiers.
Source credentials are encrypted at rest (Fernet) and decrypted only during a run. Passwords are bcrypt-hashed.
Every project's data is namespaced and isolated; two customers using the same source never share or overwrite data.
Data contracts, bad-row quarantine with reasons, column-level lineage, and a full per-run audit trail — evidence you can show a regulator.
Generated SQL runs in a read-only sandbox; outbound fetches are SSRF-guarded; secrets never appear in URLs or logs.
The platform is built for modern data-protection regimes — the EU's GDPR, California's CCPA, US HIPAA, and India's DPDP Act: automatic PII handling, purpose-scoped access, deletion on request, and a full audit trail. For RBI-regulated lenders, masking + validation + audit map directly to the 2025 Digital Lending expectations.
Found a security issue? Email contact@sluicedata.com — we'll acknowledge promptly and work with you on a fix.
Certifications (SOC 2 / ISO 27001) are on our roadmap; this page describes controls in place today.