Rasa

Definition

Rasa is an open-source conversational AI framework for building chatbots and voice assistants, combining natural language understanding (NLU) with dialogue management that developers can fully customize and self-host.

Overview

Rasa provides two core building blocks: Rasa NLU, which extracts intents and entities from user messages, and Rasa Core (dialogue management), which decides how the assistant should respond based on conversation state. Unlike many hosted chatbot builders, Rasa is designed to be self-hosted and deeply customizable, making it popular with engineering teams that need full control over data privacy and conversational logic. Developers define intents, entities, and conversation flows in configuration files, then train NLU and dialogue models on that data. Rasa can be extended with custom actions written in Python to call external APIs, databases, or business logic mid-conversation, and it supports integration with channels like websites, Slack, and messaging apps. Rasa competes with hosted platforms such as Dialogflow, Amazon Lex, and Watson Assistant, but differentiates itself through open-source flexibility and on-premises deployment options, which appeal to regulated industries like banking and healthcare. It is frequently covered alongside courses on AI Agents & Agentic Workflows.

Key Features

• Open-source natural language understanding (NLU) and dialogue management
• Self-hostable for full data control and on-premises deployment
• Custom Python actions for integrating external APIs and business logic
• Contextual dialogue management that handles multi-turn conversations
• Support for multiple channels including web, Slack, and messaging apps
• Rasa X / Rasa Pro tooling for conversation review and model improvement
• Extensible pipeline for combining rule-based and machine-learning NLU

Use Cases

Frequently Asked Questions

Related in AI Tools