About This Event
Classifying text at scale usually means an LLM call per row or a specialized model you have to train. prompt_jev() is SQL-native classification: 100,000 rows at 89% accuracy in 40 seconds. Jacob Matson, Dumky de Wilde, and Hamilton Ulmer run it live.
What You'll Learn
- See where prompt_jev() fits between per-row LLM calls and models you train yourself
- Call prompt_jev() from SQL and get labels plus confidence scores you can filter, join, and aggregate
- Read the benchmark behind the launch: 89% accuracy on 100,000 AG News rows in 40 seconds, about 2,500 rows per second
- Classify an entire table without waiting half an hour for the result
Who Should Attend
This event is ideal for:
Data EngineersData AnalystsML Engineers
