AI and Shiny for Python: Unlocking New Possibilities
summary
Shows a working Shiny for Python workflow that uses the chat_stream package to call OpenAI, returning structured JSON (recipes) from messy webpages and letting a human validate before database insert; it uses GPT-3.5-turbo-16k for larger context and a vector database to augment the model with docs, and notes deployment risk (API keys exposed on Shiny Live) and that the model only returns text so plots must be generated by running code separately.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 0:44What is a large language model
- 2:22How do you create a conversation
- 4:36How does it talk to the API
- 6:12Prompt engineering
- 7:05Different types of AI
- 11:36Validation
- 12:51Knowledge Augmentation
- 15:38Conclusion
related talks
20mBuilding ML and AI apps with Shiny for Python
36mGenerative AI + Python: Unlocking Efficiency, Personalization, and Insight
45mBringing the Power of AI to Your Application
58mPractical Tips for Using Generative AI in Data Science Workflows
36mAI for Government Efficiency with ChatGPT: From Hype to Reality
32m"How Smart is AI? – Real projects to build real understanding"
20mOld Apps, New Tricks: How AI can write Automated Tests for your Shiny Apps
46mUsing AI in Software Design: How ChatGPT Can Help With Creating a Solution Arch.