Rasa is an open source machine learning framework for automating text and voice-based conversations through natural language understanding and dialogue management.
The framework addresses the challenge of building contextual conversational assistants that can handle multi-turn interactions. It combines natural language understanding (NLU) for intent recognition with dialogue management to maintain conversation state and logic. The tool integrates with multiple messaging platforms including Slack, Facebook Messenger, Telegram, Twilio, and others, allowing developers to deploy trained models across different channels without rewriting core logic.
Rasa Open Source is currently in maintenance mode, with development focus shifting toward Hello Rasa and CALM (Conversational AI with Language Models), a newer approach that combines large language model fluency with business logic enforcement through flows rather than rigid dialogue trees. The legacy framework suits teams building traditional intent-based chatbots who need fine-grained control over conversation flows and want to avoid dependency on external LLM APIs. Developers should be aware that the classic Rasa Open Source represents an earlier architectural approach; the project recommends Hello Rasa for new projects seeking a browser-based, setup-free experience with built-in LLM capabilities.
Work in the issue tracker centers on bug fixes and enhancements to the open source framework, with activity concentrated in the core Rasa OSS area.
