Platform / Governance & Observability

Control over every action

CData gives IT and security teams policy enforcement and real-time observability over every AI-to-data interaction.

Audited against the standards you require
Trusted by enterprise teams
GSK
Palantir
Anthropic
Office Depot
Google
Fit Check

Will this actually work in my environment?

Access control, audit, and observability that plug into the identity providers and security tooling your teams already run.

Access control models

RBAC • ABAC • Time-windowed access • Custom business rules • Agent service accounts

Identity providers

Okta • Azure AD • Ping Identity • SCIM 2.0

Audit destinations

Query-level logs • SIEM-ready export • Real-time dashboard

Deployment

Cloud • Hybrid

The Problem

Most enterprise AI deployments create governance gaps fast.

AI operates outside the controls you already built

Shared service accounts let AI access whatever the account can see—not what the requesting user should see.

Unclear what AI actually accessed

Logs may show something ran—not what was queried, what was returned, or who was behind it.

Stopping a runaway agent is too blunt or too slow

Without granular controls, incident response becomes an all-or-nothing shutdown.

How It Works

One control plane for governed AI access.

Every interaction is controlled, logged, and observable — without separate governance workflows or delayed enforcement.

01

A user or AI agent sends a request through CData.

02

CData evaluates the request against access policies and user identity before it reaches the source.

03

Identity is resolved at the identity & access layer. CData then evaluates the request against governance policies before execution.

04

The interaction executes under governance, and query-level audit details are captured.

05

Real-time observability reflects the interaction immediately, including query activity, system health, and available control actions.

Control, audit, and observe — every AI interaction, in one place.

Key Capabilities

Control, visibility, and response, built in.

01
Access boundaries

AI stays inside approved access boundaries

Apply RBAC, ABAC, time-based controls, and business rules so AI requests stay within defined policy.

AI stays inside approved access boundaries
02
Real-time monitoring

Monitor production AI in real time

See and take action on query volume, success rate, connection health, and anomalous behavior from one dashboard.

You can monitor production AI in real time
03
Audit trail

Every AI interaction is auditable

Query-level logs capture who initiated the request, what ran, which agent was involved, and what data was returned.

Every AI interaction is auditable
04
Instant kill switch

You can shut down access instantly

Revoke one user, disable one connection, suspend one workspace, or lock down the full account in seconds.

You can shut down access instantly
05
Layered controls

Layer access controls on top of the source system

RBAC and ABAC rules enforced on top of source system permissions — IT can restrict AI access further without needing to touch the source system.

Layer access controls on top of the source system
06
Everywhere

The same governance model works everywhere

Apply the same controls and visibility across cloud, on-prem, and hybrid systems without governance gaps.

The same governance model works everywhere
AI stays inside approved access boundaries
You can monitor production AI in real time
Every AI interaction is auditable
You can shut down access instantly
Layer access controls on top of the source system
The same governance model works everywhere
Governance and observability in action
“Security is a real concern with the current MCP server marketplace. We want to leverage OAuth specifically—so users are properly identified and the permissions we've established for them extend seamlessly into AI functionality. CData can do that.”
Nick Bisciotti
Director of IT, HighRes BioSolutions
Implementation Path

Implement your policies and track adherence

Day 01

Govern & log

Connect the first data source and activate passthrough identity and baseline audit logging.

Milestone: AI queries are governed by source-system permissions, with first audit activity visible immediately.

Day 03

Policy & SIEM

Add policy controls and connect audit logs into your security workflow.

Milestone: Fine-grained access controls are active, and audit logs are flowing into existing security operations processes.

Day 07

Observe & test

Configure observability thresholds, policy alerts, and incident-response testing. (Identity lifecycle is configured on Day 3 of the Identity & Access rollout path.)

Milestone: Identity updates are automated, observability is live, and incident-response controls are tested and operational.

Security & compliance

Governed from the first query to the final action.


Governance model
  • No data movement — data queried in place; no intermediate copies created.
  • Zero-trust compatible — passthrough identity enforces least privilege by default; no standing access for AI agents.
  • SIEM-ready audit logs — export to Splunk, Datadog, or your security platform of choice.
  • GDPR support — data minimization, in-place access, comprehensive audit capabilities.
  • PII detection and control — configurable warn, redact, or block policies enforced at the MCP/API tool-call boundary—inbound and outbound
  • Enterprise key vaults — credentials stay in your own infrastructure under your own access controls and rotation schedules
  • Centralized audit logging — immutable, per-account audit storage with a full-text searchable API and no record cap
Certifications & security
  • SOC 2 Type II — Completed.
  • ISO/IEC 27001:2022 — Completed.
  • Third-party pen testing — conducted regularly; independent findings available under NDA.
  • AES-256 at rest — exceeds Federal compliance minimums.
  • TLS 1.3 in transit — all data in transit protected.
FAQ

Questions teams ask first.

Get AI under control before it becomes a governance problem.

Talk to our team about access controls, audit requirements, and observability for your environment. Or explore how Governance & Observability fits into the broader CData platform.