Introduction

Erato is a sovereign, open-source AI-backed chat interface designed for organizations that prioritize data sovereignty and security. Built with enterprise needs in mind, Erato brings powerful AI capabilities directly into your existing IT infrastructure while ensuring your data never leaves your control.

What is Erato?

Erato is an open-source chat UI for Large Language Models (LLMs) that serves as a comprehensive AI-backed chat interface for organizations. Unlike traditional SaaS solutions, Erato is designed from the ground up for self-hosting, enabling you to deploy AI capabilities on-premises or in your private cloud environment.

At its core, Erato addresses the critical need for data sovereignty in AI deployments while providing enterprise-grade features such as Single Sign-On (SSO) integration, advanced access control, and extensible architecture through the Model Context Protocol (MCP).

Core Approaches and Philosophy

Privacy-First and Data Sovereignty

Erato’s fundamental principle is that your data stays in your house. Every component is designed to operate within your infrastructure boundaries:

• On-premises deployment: Full control over where your data is processed and stored
• Private cloud compatibility: Deploy in your own cloud environment with complete isolation
• No external data transmission: All AI processing happens within your controlled environment
• Compliance-ready: Built to meet strict European data protection requirements

Enterprise-Native Identity Integration

Erato is SSO-native and supports a wide range of identity providers, making it seamlessly integrate with existing corporate infrastructure:

• Microsoft Entra ID (Azure AD) with first-class support
• OAuth2/OIDC compatibility with providers like Google, Keycloak, and others
• Resource-based access control that respects your existing permission structures
• Seamless user experience with corporate login credentials

3. Open Source and Vendor Lock-in Freedom

Built with transparency and freedom in mind:

• Fully open source under AGPL-3.0 license
• No vendor lock-in: You own and control the entire stack and underlying data
• European alternative: Reducing dependency on US-based AI platforms

4. Extensible and Future-Proof Architecture

Erato’s architecture is built for adaptability and growth:

• Model Context Protocol (MCP) support: Connect to external data sources, APIs, and tools
• Multi-model compatibility: Work with various LLM providers (OpenAI, Azure OpenAI, local models)
• Kubernetes-native deployment: Production-ready with Helm charts included
• Small resource footprint: Efficient deployment suitable for organizations of all sizes

Target Use Cases

Erato excels in scenarios where organizations need:

• Document-heavy knowledge work with secure AI assistance
• Integration with existing enterprise systems (M365, SharePoint, internal APIs)
• Compliance with strict data protection requirements
• Data-integrated AI approach through MCP server integrations
• Organizational AI strategy that builds on open standards

Whether you’re an organization looking to implement AI securely or seeking a sovereign alternative to US-based AI platforms, Erato provides the foundation for a comprehensive, controlled, and compliant AI strategy.