Rasa

Rasa provides an open source framework for building conversational assistants that handle contextual, multi-turn dialogue over text and voice. It combines trainable natural language understanding for intent classification and entity extraction with machine-learning-driven dialogue management, plus custom actions, forms, and slot filling for gathering information across turns. Connectors ship for Facebook Messenger, Slack, Telegram, Twilio, Microsoft Bot Framework, Rocket.Chat, and custom channels, and the stack self-hosts so conversation data stays under the operator's control. The classic framework is now in maintenance mode while Rasa Technologies focuses on its CALM approach, which pairs LLM flexibility with strict business logic in flow-based assistants, and on newer hosted tooling. Released under the Apache 2.0 license, Rasa has been used by developers and enterprises to run production chatbots and voice assistants where data control and predictable behavior matter.