EvolutionAPI/evolution-api
Evolution Api
Evolution API is a collection of system components including a WhatsApp API gateway, a multi-channel messaging bridge, and a conversational AI orchestrator. It functions as an event-driven messaging middleware that links messaging platforms with large language models and external applications to automate text and audio responses.
The project provides a self-hosted marketing automation platform for executing customer relationship workflows and outreach campaigns. It further distinguishes itself by routing chat conversations between different messaging services and customer support tools through a unified interface.
The system covers omnichannel messaging management, real-time event streaming to external queues, and the automation of business communications. It also includes capabilities for syncing conversations with bot builders and offloading binary media to cloud object storage.
Features
- Conversational AI Platforms - Implements a coordination layer that links messaging platforms with large language models to automate conversational responses.
- Messaging API Interfaces - Provides a single abstraction interface to control and automate interactions across multiple different messaging platforms.
- AI Agent Orchestration - Provides a coordination layer for managing specialized AI agents to execute complex operational and development tasks.
- AI Model Integrations - Integrates messaging flows with language models to automate text responses and convert audio to written text.
- Chatbot Integrations - Links messaging platforms with language models to automate text responses and audio-to-text conversion.
- Marketing Automation Platforms - Provides a self-hosted management platform for executing customer outreach campaigns and engagement workflows.
- Messaging Campaign Platforms - Manages customer relationship workflows and communication sequences via a self-hosted messaging campaign platform.
- WhatsApp Messaging Integrations - Enables linking messaging flows to WhatsApp Business APIs to send and receive communications at scale.
- API Gateways - Ships a server that provides a REST interface to automate WhatsApp messaging and integrate it with external applications.
- Chat Platform Integrations - Synchronizes user conversations with external customer service tools and bot builders to manage interactions.
- Message Bridges - Provides a unified interface for routing and transporting chat conversations between different messaging services and support tools.
- Messaging Adapters - Unifies diverse messaging protocol implementations behind a single API interface for consistent request and response structures.
- Messaging Orchestrators - Coordinates and routes message delivery across multiple communication channels from one central interface.
- Event Middleware - Provides a communication layer that streams real-time chat events to external queues and storage for asynchronous processing.
- Customer Support Routing - Syncs chat history and active conversations with external customer service tools for team-based interaction management.
- Event Streaming Platforms - Pushes real-time system events to external queues or streaming platforms for asynchronous processing.
- Event-Driven Webhooks - Pushes real-time application events to external endpoints using asynchronous HTTP webhooks for decoupled processing.
- Messaging Session Persistence - Maintains persistent WebSocket connections to messaging clients to synchronize real-time message states.
- AI Model Middleware - Intercepts incoming and outgoing message flows to inject processing steps from external large language model providers.
- API Gateways - Handles incoming requests and routes them to the appropriate session instance via a stateless API gateway.
- Communication Integrations - Open-source WhatsApp integration API.
Star history
How this analysis was created: This summary and feature list were written by an AI model that read the project's README and public documentation pages. Each feature links to the documentation it came from; stars, license and language come straight from the GitHub API. The model does not read the source code, and the analysis is refreshed when the project is re-analysed. Learn more on our About page.
