Rasa

Rasa is an open, extensible platform for building production-grade AI assistants and conversational agents. It provides developers with tools to design, train, and deploy context-aware assistants that work across chat, voice, and custom channels. At its core, Rasa combines natural language understanding (NLU) with dialogue management to interpret user messages, track conversation state, and select appropriate responses or actions.

Key capabilities include intent classification, entity extraction, and customizable policies for conversation flows. Rasa supports retrieval-augmented generation (RAG), tool calling, and integration with external APIs and backends, enabling assistants to perform tasks such as booking, troubleshooting, account management, or knowledge retrieval. Its modular architecture allows teams to plug in custom models, connect to LLMs, and manage complex business logic.