What it does Rasa UI is an Angular/Node.js frontend for the Rasa conversational AI framework. It gives you browser-based forms to create intents, entities, stories, actions, and responses—the raw material that feeds Rasa’s NLU and Core pipelines. You can also train models, swap between versions, and log API requests for later analysis.

The interesting bit The project decouples data management from the runtime: you can use it purely as a training-data editor, storing everything in SQLite, then hand the files off elsewhere. That’s useful if your Rasa instance lives somewhere else, or if you’re just prototyping.

Key highlights

• Supports multiple bots from a single instance
• Model versioning for A/B testing or rollback
• Request logging by proxying Rasa API calls through the UI (/api/v2/rasa/...)
• Docker and docker-compose setups included
• SQLite backend for easy backup and sharing

Caveats

• Upgrading from pre-v3.0.0 is a hard cut: the project switched from Postgres to SQLite with no migration path
• Database schema changes between versions require manual migration script runs
• Developed against Rasa 1.2+; compatibility with newer Rasa versions is unclear from the README

Verdict Worth a look if you want a visual training-data editor for Rasa or need to manage multiple bot projects in one place. Skip it if you’re already happy with YAML and version control, or if you need guaranteed compatibility with the latest Rasa releases.