rasa-ui

Why we included this project

Building assistants on Rasa usually means editing YAML training files by hand and retraining from the command line, which gets cumbersome once you have more than a few dozen intents. Rasa UI replaces most of that with a browser-based interface: you can create and edit intents, entities, synonyms, regex patterns, stories, actions, and responses, then train and load different model versions without leaving the page. A single instance can manage several bots at once, and because it can run without a live Rasa backend, it also doubles as a standalone training-data editor for teams that want to prep data before connecting a server. It logs API requests too, so you keep a record of usage and test history to improve models later. It's a pragmatic fit for teams that prefer a visual workflow over raw config files, even if it means running one more self-hosted service.