Data Creation: How Organizations Are Deliberately Creating Data To Power ML & Data Applications
summary
Argues organizations should stop relying only on extracted sources and intentionally create data for ML and products; gives two concrete wins: an insurer buys satellite imagery, measures a few hundred property perimeters, trains an ML model to infer perimeters across tens of thousands of policies to model flood exposure, and a bank captures form-fill timings and finds that this behavioral feature is 4x more predictive for fraud. Describes building behavioral pipelines with SDKs and partners like Databricks and dbt.
AI-written summary · about
chapters (7)
- 0:00Introduction to data creation
- 1:48Limitations of data as oil metaphor
- 5:18Challenges of extracted data
- 14:24Why deliberately create data
- 19:42Global data creation trends
- 20:45Snowplow's role in data creation
- 22:21Q&A on data creation
related talks
18mBehavioral Data Creation for AI - Delivering Value and Insights
45mInnovating with Data
39mDoing Data: The Critical Process of Data Preparation (...)
26mUnlocking the Power of Data-Centric AI.
36mArchitecting Data: A Deep Dive Into the world of Synthetic Data
32mMaking Models Work: How an optimized ML lifecycle can drive better ROI
29mBecoming a data-driven enterprise
01h 01mYou don't need AI; You need to know what your data is.