Security & Compliance

AI usage your security team can approve

Centralize security and privacy across AI usage with sensitive-data redaction, request logs, audit logs, SSO, RBAC, and ZDR options.

Redaction pipeline

Show what entered the system, what was redacted, and what was logged for review.

Identity

SSO

Access

RBAC

Data

ZDR + PII

Request

PII detected

Card and address entities were redacted.

Actor

Support app

Traffic tied to a scoped Universal API key.

Audit

Policy changed

Admin update tracked with time and actor.

New capabilities

What your team gains with Concentrate

01

Control who uses what

Map people, teams, keys, roles, and model access.

02

Reduce data risk

Use ZDR options and PII redaction where sensitive traffic needs protection.

03

Keep audit records

Track request logs and settings changes for internal reviews. Request our SOC 2 Type II report from the Trust Center.

Where it fits

For GRC, InfoSec, legal, and internal audit

Security teams need proof: who sent the request, which key owned it, what data was filtered, and which policy applied. For patient data, the HIPAA LLM Gateway is $499 a month and gives you access to all of our BAAs. API and inference details are on HIPAA Inference. Request the SOC 2 Type II report from the Trust Center.

Healthcare and BAA-covered inference

You sign one BAA with Concentrate for covered inference. We maintain separate BAAs with the providers serving the approved routes. Pricing is on HIPAA Inference. Activation locks zero data retention on, locks request-content logging off, and limits traffic to covered provider routes. Those locks are on the HIPAA LLM Gateway. The covered routes are the BAA-eligible filter in Model Fortress.

Sensitive data handling

Redact PII, PCI, and PHI from prompts, responses, or both before traffic reaches a model provider.

Access review

Use SSO, RBAC, team workspaces, and API keys to limit who can use what.

Audit evidence

Keep request logs and audit logs for settings changes, access changes, and model usage reviews.

SOC 2 Type II

Concentrate has a completed SOC 2 Type II report. Review our program and request the report under NDA in the Trust Center.

Provider policies

Turn zero data retention on and enforce it by provider, team, or key where policy requires it.

Security & Compliance basics

Frequently asked questions

What evidence does a security team need for AI usage review?
Security usually needs to know who can call models, which keys map to which apps, what data can be sent, whether PII redaction or ZDR options apply, and whether request logs and audit logs exist for review. For platform assurance, request our SOC 2 Type II report from the Trust Center.
Can model access be restricted by team or app?
Yes. Use roles, workspaces, API keys, and model access settings to limit which teams or apps can call specific models or providers. A customer-facing app can have stricter access than an internal evaluation project.
How should sensitive prompt data be handled?
Use PII redaction where sensitive fields may appear, enable ZDR options where provider retention is a concern, and keep logs scoped so reviewers can see redaction events without exposing raw secrets unnecessarily.