Welcome to OPAQUE¶

OPAQUE is a confidential AI platform for running agentic workflows and analytics jobs on sensitive data without exposing it during processing. Work happens inside protected workspaces with built-in access controls and auditability, so teams can build and operate AI systems over private or governed data.

How to use this guide¶

If you’re new to OPAQUE, begin with Get started. It introduces the UI and walks you through the initial setup and first steps for working in a workspace.

For a conceptual overview, see What is OPAQUE to understand how the platform is organized, the types of workloads it supports, and how governance works.

Once you’re familiar with the basics, you’ll typically focus on one of the following areas:

• Agentic workflows: Build long-running, callable services that combine models, retrieval, and structured logic. Launch a workflow once, then trigger it via the OPAQUE SDK or API.
• Analytics workflows: Run batch jobs over protected datasets using Python (PySpark) or SQL. Jobs execute to completion and are designed for analysis, transformations, and model development.

Who this guide is for¶

This guide is for anyone using OPAQUE to build and run AI workloads on sensitive data, especially in environments with privacy, security, or compliance requirements.