Rasa
Open-source framework for building contextual AI assistants
About
Rasa is the open-source machine learning framework for building conversational AI — chatbots and voice assistants where the team owns the whole stack. It combines NLU (intent classification, entity extraction) with ML-based dialogue management, customizable pipelines, and multi-language support, deployable on your own infrastructure for data-sensitive support use cases. The newer CALM (Conversational AI with Language Models) engine pairs LLM fluency with deterministic business logic: the model understands the user, code enforces the rules. Installation is a pip install away, and the GitHub repo is among the most starred in conversational AI. The tradeoff is that it's developer infrastructure, not a product: non-technical teams struggle, you own hosting, training, and ML ops, and the commercial Rasa Platform exists for teams that want managed scale.
What it can do
- Open-source NLU and dialogue management; CALM engine; custom pipelines; multi-language support; omnichannel connectors; on-premise deployment; LLM integration.
Setup
pip install rasa, rasa init to scaffold a project, define NLU data and flows, train, connect to channels.
Prerequisites
Python 3.7+; ML/NLP engineering capacity; infrastructure for hosting.
Use cases
Works with
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Open-source framework (Apache 2.0); NLU + dialogue management architecture verified via Medium guides and selecthub; CALM engine (LLM + deterministic business logic) documented in GitHub README; Among most-starred conversational AI repos; pip-installable, on-premise deployable; Limitation: developer infrastructure, not a product; team owns hosting, training, and ML ops.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.