Using GenAI with Audio Data for Driving Better Customer Experience
summary
Shows a concrete pipeline: audio files run through Whisper on Snowpark Container Services to produce transcripts, custom prompts extract tabular fields, vectors are stored inside Snowflake and a RAG layer answers queries, and a fine‑tuned text‑to‑SQL model generates executable SQL. Uses Cortex for LLMs-as-SQL-functions, Streamlit for the UI, and Nvidia GPUs for model hosting. A live customer application (claims/policy joins, sentiment, first‑call resolution) returns near‑perfect answers in demos and is already running in some customer flows.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 1:47Snowflake Data Cloud overview
- 7:33GenAI features on Snowflake
- 13:15Snowflake Cortex and SPCS
- 14:17Insurance call center use case
- 21:20Customer Insight Analyzer demo
- 26:33Q&A and closing
related talks
29mBuilding End to End Gen AI Powered Analytics: From Data to Production
22mDoing Better with Data and AI
36mImplementing Natural Language Analytics with Snowflake Cortex
30mBuilding Data AI Application with the Snowflake AI Data Cloud
17mPrepare to master common generative AI use cases with ease
23mContact Center Voice AI: Low-Latency Intelligence Extraction from Messy Audio Streams
01h 13mBuilding a Semantic AI Layer in Snowflake: From Metrics to AI Insights
26mEnabling the next generation in voice technology