Building End to End Gen AI Powered Analytics: From Data to Production
summary
Snowflake demonstrates an end-to-end GenAI pipeline that ingests S3 files and notebooks, runs SQL and Python with cortex functions and a model garden, and processes 118,000 reviews to translate, categorize (awful → excellent), compute sentiment on a -1 to 1 sentiment scale, and aggregate by truck. The demo shows no API or GPU management, batch-optimized LLM calls, optional embeddings/vector DB for RAG, Iceberg table interoperability, and a reproducible notebook that surfaces actionable issues (e.g., Truck 5: cold food, poor service).
AI-written summary · about
chapters (15)
- 0:00Introduction to Snowflake
- 1:50Snowflake’s AI use cases
- 3:38Workflow on Snowflake platform
- 5:42LLM functions in Snowflake
- 9:56Activate your data with AI
- 11:21Demo setup and data overview
- 12:25Sentiment analysis and translation
- 15:53Model comparison and prompt tuning
- 18:01Generating business insights
- 20:28Analyzing customer feedback
- 21:52Visualizing sentiment data
- 23:15Generating recommendations
- 24:41Building integrated pipelines
- 25:45Empowering end-to-end visibility
- 27:32Q&A session
related talks
23mAcceleratinag Data Science and Machine Learning with Snowflake DataTech
01h 13mBuilding a Semantic AI Layer in Snowflake: From Metrics to AI Insights
17mPrepare to master common generative AI use cases with ease
30mBuilding Data AI Application with the Snowflake AI Data Cloud
31mAI-driven DataOps: Building data products with a taste of the future
27mUsing GenAI with Audio Data for Driving Better Customer Experience
15mEnterprise AI Runs on Trusted Data
32mScalable Python and SQL Data Engineering without Migraines