Lenny Blyuhker - BigDataSV 2014 - theCUBE
summary
JiWire describes building audience profiles from billions of location tags with a patented Location Graph and a mobile ROI product, Location Conversion Index, that matches device IDs to store visits to measure incremental traffic. Pipeline uses Hadoop and Hive for ETL, shifts to columnar databases and ParAccel for fast analytics, and deploys in-memory layers to meet 50 milliseconds bidding SLAs while separating exploratory and reporting workloads for stability.
AI-written summary · about
chapters (12)
- 0:00Introduction to Big Data SV
- 1:16Evolution of big data focus
- 2:40Jiwire's business and location graph
- 4:03Data vs information in IoT
- 5:05Tech evolution for analytics
- 7:36Location graph product details
- 8:39Location conversion index
- 9:42Real-time vs batch analytics
- 11:27High performance analytics trends
- 13:13ParAccel technology benefits
- 14:38Actian partnership insights
- 15:43Business value in big data
