What is Eden?

Eden Substrate is a governed operating layer for AI, data, modernization, operations, and security across local, private, and public environments.

Customers discover Eden through five product areas:

The underlying Eden products implement those areas:

The Problem

Modern applications and AI agents need to work across many infrastructure surfaces:

• PostgreSQL for transactional data
• MongoDB for document storage
• Redis for caching and sessions
• Pinecone for vector search
• ClickHouse for analytics
• LLM providers, applications, APIs, platforms, and agents for governed AI and operational workflows

Managing each surface separately requires:

• Learning different protocols and query languages
• Integrating multiple client libraries
• Handling different connection patterns
• Managing credentials and permissions for each system
• Monitoring each system separately
• Coordinating migrations, policy, and rollback across systems

This complexity slows down development and increases operational overhead.

The Solution

Eden provides a governed control plane for infrastructure access. Teams use Eden to:

• Put Gateway in front of models, tools, APIs, and databases
• Give people and agents governed access through AI Orchestration
• Move live database workloads through Live Migrations without taking applications offline
• Apply policy, telemetry, approvals, and audit trails consistently across products

Start with Products Overview when you need the product map before implementation details.

Core Concepts

Organizations

An organization is a tenant in Gateway. Each organization has:

• Its own users and administrators
• Its own endpoints (database connections)
• Isolated data and permissions
• Independent access control

Organizations enable multi-tenancy, allowing a single Gateway instance to serve multiple teams or customers.

Endpoints

An endpoint is a connected database or service. Each endpoint has:

• A unique name within the organization
• Connection configuration (URL, credentials, etc.)
• A type (postgres, mongo, redis, etc.)
• Access permissions

Think of endpoints as named database connections that you can query through the API.

Authentication

Gateway supports multiple authentication methods:

• Basic Auth - Username and password
• Bearer Tokens - JWT tokens for programmatic access
• API Keys - Organization-scoped keys for services

All requests are scoped to an organization based on the authenticated user.

Role-Based Access Control (RBAC)

Control who can access what with fine-grained permissions:

• Subjects - Users and roles
• Resources - Endpoints, workflows, templates
• Actions - Read, Write, Delete, Execute, Admin

Example: A "data_analyst" role can Read from analytics endpoints but cannot Write or Delete.

Workflows

Workflows are sequences of operations across multiple endpoints that can be executed as a unit. Benefits:

• Coordinate operations across databases
• Reuse common patterns
• Atomic execution with rollback support
• Template-based for consistency

Key Features

1. Multi-System Support

Eden currently supports:

2. Governed Access

Route requests through one policy and telemetry path instead of handing every application direct credentials to every system. Eden can enforce permissions, rate limits, audit logging, and endpoint-specific controls at the gateway.

3. Cross-Database Transactions

Gateway can execute ACID transactions across different database types:

{
  "transaction": {
    "operations": [
      {
        "endpoint": "postgres_db",
        "query": "INSERT INTO orders ..."
      },
      {
        "endpoint": "mongo_db",
        "operation": "updateOne",
        "collection": "inventory"
      },
      {
        "endpoint": "redis_cache",
        "command": "SET",
        "key": "order:123"
      }
    ]
  }
}

If any operation fails, all changes are rolled back.

4. Built-in Observability

Every request includes:

• Distributed Tracing - OpenTelemetry traces across all operations
• Metrics - Request latency, error rates, database performance
• Structured Logging - Context-rich logs with trace correlation
• Health Checks - Monitor endpoint availability

5. Horizontal Scalability

Gateway scales horizontally:

• Load Balancer - Distribute requests across multiple nodes
• Single or HA Profiles - Run one gateway or an active-active/active-passive topology
• Embedded Runtime State - Internal caching uses embedded ShardKV/ShardMap
• No Internal Redis Dependency - Redis or Valkey is needed only when selected as a data endpoint
• Connection Pooling - Efficient database connection management

6. Developer Friendly

• REST API - Use from any programming language
• JSON Everywhere - Simple, familiar data format
• Comprehensive Docs - Examples for every feature
• Local Development - Docker Compose for easy setup

Architecture Principles

Eden is built on several key principles:

API-First Design

Everything is accessible through a well-documented REST API. No proprietary protocols or SDKs required (though client libraries are available).

Infrastructure Agnostic

Eden does not favor one database, model provider, API, or tool surface. Each endpoint type is a first-class resource with policy, telemetry, and access controls.

Security by Default

• All communications can use TLS
• Authentication required for all operations
• Fine-grained RBAC for access control
• Audit logging for compliance

Cloud Native

• Containerized deployment
• Kubernetes-ready
• Horizontal scaling
• Graceful degradation

Use Cases

1. Polyglot Persistence

Your application needs different databases for different purposes. Gateway provides a unified interface while letting you use the right tool for each job.

2. Database Migration

Migrating from one database to another? Use Live Migrations to proxy traffic, mirror writes, validate behavior, roll back safely, and gradually shift traffic.

3. Multi-Tenant SaaS

Each customer is an organization with their own endpoints and data. Gateway handles the isolation and access control.

4. Gateway for Infrastructure

Expose models, databases, APIs, and tools through a controlled path with authentication, rate limiting, policy, and monitoring.

5. Data Integration

Coordinate data across multiple systems with workflows and cross-database transactions.

Next Steps

Ready to get started?

1. Quick Start - Get Gateway running in 5 minutes
2. Basic Concepts - Deeper dive into core concepts
3. First Steps - Your first organization and endpoint

Learn More

• Introduction
• API Reference
• Examples