Going Beyond “Sorry, I Didn’t Get That”: Building AI Assistants that Scale...

on YouTube ↗01h 29m

" summary

Practical workshop demonstrating how to build scalable ML-only conversational assistants using Rasa NLU and Rasa Core. Covers intent classification, entity extraction/slot filling, dialogue policies, response selection with confidence scores, connectors to messaging platforms, an interactive learning loop that ingests real conversations and feedback, plus a training pipeline and notes on running the models in production with examples of what worked and what required iteration.

AI-written summary · about

" chapters (12)

1. 0:00Scalable AI assistants
2. 3:17Open source frameworks
3. 11:53Handling user input
4. 20:36Intent classification
5. 29:56Live demo
6. 36:21Model improvements
7. 44:00Scaling to millions
8. 55:48Robust techniques
9. 1:06:22Domain customization
10. 1:15:46Interactive learning
11. 1:22:17Deployment challenges
12. 1:27:20Q&A

" related talks

• 01h 31mGoing beyond “Sorry, I didn’t get that” building AI assistants that scale...Justina Petraitytė · PyData DC 2018
• 01h 38mConversational AI with Rasa Core & NLUTom Bocklisch · PyData Amsterdam 2018
• 30mConversational AI: Building clever chatbotsTom Bocklisch · PyData Berlin 2017
• 01h 27mDeprecating the state machine: building conversational AI with the Rasa stackJustina Petraitytė · PyData Berlin 2018
• 47mBuilding a cloud AutoML platform for creating voice robotsArtem Bondar · Saint HighLoad++ 2022
• 33mBuilding the next-generation Conversational AI with Python and Deep LearningDavid Low Jia Wei · PyCon AU 2017
• 31mRasa: open source conversational AI to build next generation chatbotsJoey Faulkner · PyCon.DE 2017
• 46mFullstack AI: The Building Blocks for Production Ready AI-Powered SoftwareSebastian Nilsson · AI-fokus 2026