Driving the On-Demand Economy with Spark and Predictive Analytics
Download as PPTX, PDF•1 like•741 views
The document discusses how data scientists need real-time analytics capabilities to power the on-demand economy. It introduces MemSQL 5 as a database platform for real-time analytics that can help overcome barriers like slow loading, queries, and ongoing data processing faced with batch processing. MemSQL 5 includes features like Streamliner for building real-time data pipelines and predictive analytics using Spark and MLlib to power applications like predictive scoring and IoT.
Driving the On-Demand Economy with Spark and Predictive Analytics
- 1. Driving the On-Demand Economy with Spark and Predictive Analytics Eric Frenkiel, CEO and Co-Founder March 2016 San Jose
- 2. We live in an on-demand economy
- 4. Data scientists need to carry the world forward
- 5. Too often stuck with batch processing and slow ETL
- 6. Slow Loading Slow Queries No Concurrency On-going Data Processing Barriers
- 7. There is a shift underway… the move to real time
- 8. Introducing MemSQL 5 The Database Platform for Real-Time Analytics
- 14. Hadoop Apache Spark STREAMLINER Extending Hadoop to Real-Time Processing
- 15. Spark MLlib Predictive Model Raw Sensor 1 + Predictive Score 1 Real-Time Scoring for Predictive Applications ID S1 S2 P1 P2 1 … Real- Time Inputs Apache Spark STREAMLINER Business Logic
- 16. Putting Streamliner to work for IoT applications
- 20. Real-TimeBatch
- 22. TUTORIAL 11:00AM ROOM 210 DASH FORWARD From Descriptive to Predictive Analytics Guest from Kellogg Be sure to visit the MemSQL booth