Deprecating the state machine: building conversational AI with the Rasa stack
summary
Live code shows a concrete pipeline: Rasa NLU & Rasa Core plus SpaCy vectors, SVM intent classifier and conditional random fields for entity extraction, a tiny dataset (roughly seven stories, a few dozen NLU examples) and a domain with slots, templates and a custom action that calls a public image API. The system trains (200 epochs in the demo), supports interactive learning and online correction, and the chatbot runs — the image-retrieval demo is built but fails live due to network/API issues.
AI-written summary · about
chapters (14)
- 0:00Conversational AI with Rasa
- 6:21Rasa vs hosted solutions
- 8:13Roadmap and setup
- 9:59Rasa NLU intent/entities
- 17:33NLU model workshop
- 23:47NLU training data
- 30:18Entity extraction in Rasa NLU
- 44:41Dialogue management with ML
- 53:06Building dialogue model
- 1:01:15Domain and custom actions
- 1:11:22Training dialogue model
- 1:15:47Testing chatbot
- 1:20:24Interactive learning updates
- 1:24:01Final thoughts
related talks
01h 38mConversational AI with Rasa Core & NLU
30mConversational AI: Building clever chatbots
31mRasa: open source conversational AI to build next generation chatbots
01h 31mGoing beyond “Sorry, I didn’t get that” building AI assistants that scale...
01h 29mGoing Beyond “Sorry, I Didn’t Get That”: Building AI Assistants that Scale...
11mTalking to my past self (without introducing temporal paradoxes!)
34mWriting a Chatbot in PHP
46mFullstack AI: The Building Blocks for Production Ready AI-Powered Software