Skip to content

Welcome to OPAQUE

OPAQUE is a confidential AI platform for building, deploying, and verifying agentic and analytics workflows over sensitive or regulated data. Built on confidential computing, OPAQUE keeps data protected at every stage—when stored, in transit, and while being processed—so even OPAQUE cannot access the underlying data. This enables teams to safely develop and run AI systems that operate on private or governed datasets

What you can do with OPAQUE

With OPAQUE, you can:

  • Build agentic workflows: Create long-running, secure AI workflows that combine large language models (LLMs), retrieval systems, and structured logic. Agentic workflows stay active after launch, allowing you to interact with them programmatically through the OPAQUE SDK or API.
  • Deploy RAG-powered services: Ground LLM responses in private data without moving or revealing it, ensuring every query and response respects your organization’s policies.
  • Run analytics workflows and jobs: Analyze and transform encrypted datasets using Python or Spark. Analytics workflows run as discrete jobs that execute once in a trusted enclave—ideal for data exploration, model training, and secure batch processing.
  • Enforce data policies automatically: Ensure that every workflow and job complies with organizational and regulatory requirements through cryptographically verifiable audit logs.

Who should use OPAQUE

OPAQUE is built for teams that need to use AI in privacy-sensitive or regulated environments:

  • AI and ML engineers developing agentic or retrieval-augmented (RAG) workflows that rely on private data.
  • Data scientists running analytics and machine learning jobs on sensitive datasets without losing access to key insights.
  • Security and compliance teams verifying that all AI and data operations adhere to policy and privacy standards.
  • Enterprise and public-sector leaders implementing AI in regulated domains such as finance, healthcare, and government.