Successfully reported this slideshow.
Your SlideShare is downloading. ×

Making your Analytics Investment Pay Off - StampedeCon 2012

Ad
Ad
Ad
Ad
Ad
Ad
Ad
Ad
Ad
Ad
Ad
Loading in …3
×

Check these out next

1 of 20 Ad

Making your Analytics Investment Pay Off - StampedeCon 2012

At StampedeCon 2012 in St. Louis, Bill Eldredge of Nokia presents: At Nokia, we expect to save millions on avoided license fees this year on a single “Big Data” project by creating a symbiotic relationship between our traditional RDBMS storage and our newer Hadoop cluster. Our hybrid approach to data enables us to manage the convergence of structured and unstructured data, and save money. In our case we use Hadoop to process and import data into traditional systems. We have found that this use of Hadoop as a preprocessing engine has enabled maximum value to be derived from our systems, our data and our people.

At StampedeCon 2012 in St. Louis, Bill Eldredge of Nokia presents: At Nokia, we expect to save millions on avoided license fees this year on a single “Big Data” project by creating a symbiotic relationship between our traditional RDBMS storage and our newer Hadoop cluster. Our hybrid approach to data enables us to manage the convergence of structured and unstructured data, and save money. In our case we use Hadoop to process and import data into traditional systems. We have found that this use of Hadoop as a preprocessing engine has enabled maximum value to be derived from our systems, our data and our people.

Advertisement
Advertisement

More Related Content

Slideshows for you (18)

Advertisement

Similar to Making your Analytics Investment Pay Off - StampedeCon 2012 (20)

More from StampedeCon (20)

Advertisement

Recently uploaded (20)

Making your Analytics Investment Pay Off - StampedeCon 2012

  1. 1. Making Your Analytics Click to edit Master title style Investment Pay Off Bill Eldredge Sr. Manager, Nokia Data Asset August 1, 2012 StampedeCon 1
  2. 2. Overview •  Nokia and Our Big Data Challenge •  Making Hadoop Pay •  Cost Avoidance •  Product Innovation 2
  3. 3. Nokia’s History: 1865 to Now 3
  4. 4. The Four Cardinal W’s Location and Time Add Human Context to Social and Search WHAT WHO WHERE WHEN 4
  5. 5. Great Mobile Products That Sense the World CREATE A LEADING “WHERE” PLATFORM WIN IN SMART CONNECT THE INVEST IN FUTURE DEVICES NEXT BILLION DISRUPTIONS 5
  6. 6. Taking Our World Class Map Making Capabilities to the Next Level … 2.4 1500+ 400+ Validations Million Map Changes Attributes per Day 11B+ 181 Probe per Local Month Field Offices 245 Field Cars 196 Driving the Roads Countries Automatic 37k+ km Feature Roads Recognition 9 out of 10 in-car nav systems use Nokia’s Location 6 Platform
  7. 7. One Platform, Enabling Contextually Rich Mobile Experiences Content Platform Maps Positions Places Directions Guidance Traffic Smart Data Apps 7
  8. 8. Data, Data, Data +1B Search Queries Annually 11B Probe Points Processed Monthly 147M 24M Map Tiles Route Requests Served Daily per Month +19M Geocoder Requests Daily 55M Positioning Requests Daily 8
  9. 9. Data Silos 9
  10. 10. Central Big Data Analytics Platform? "   One cluster for all needs? "   Do it ourselves, or with help? "   If we build it, will they come? "   Will it pay off? 10
  11. 11. Phased Approach Oozie Map Reduce Pig Hive Quartz HBase Scribe HDFS HBase FTP HBase 2010 2011 2012 Build Enhance, Make It Robust… Get Data, and Make It Pay Use it Ourselves 11
  12. 12. •  Nokia and Our Big Data Challenge •  Making Hadoop Pay •  Cost Avoidance •  Product Innovation 12
  13. 13. Best of Both Worlds •  Key device datasets •  Key location datasets •  Dedicated reporting team •  Reporting largely self-serve •  Expensive storage (€20k/TB) •  Inexpensive storage (€2k/ TB) Key Integration Principle Aggregated Event-level Data Data Projected Savings 2012: €6 Million 13 Nokia Internal Use Only
  14. 14. Integrated Solution Touch-points/Devices Consumer-Facing/ Online Serving Front-End (“Fast World”) Enterprise-Facing/ Back-End (“Slow-World”) Oracle/MySQL Reporting Activity Hadoop Cluster Hadoop Tools Logging Cluster (Cognos, Tableau) Teradata Advanced Analysis Other RDBMS Tools Streaming Data Bulk Data Sources Sources Analytics Platform Other systems 14 Nokia Internal Use Only
  15. 15. Integrated Solution – Applied to Music On-device app usage Touch-points/Devices Consumer-Facing/ Online Serving Front-End (“Fast World”) Enterprise-Facing/ Back-End Improved Music (“Slow-World”) Recommendations Oracle/MySQL Reporting Activity Hadoop Cluster Hadoop Tools Logging Cluster (Cognos, Tableau) Teradata Advanced Analysis Other RDBMS Tools Streaming Download CDN Bulk Data Music Data logs & Logs Sources metadata Sources Collections Projected Benefit: Analytics Platform Other systems 3M Euros + 15 Nokia Internal Use Only
  16. 16. Overview •  Nokia and Our Big Data Challenge •  Making Hadoop Pay •  Cost Avoidance •  Product Innovation 16
  17. 17. The Map Is Changing… Cities are possibly our 60% greatest World’s Population in Cities in 2030 achievement, constantly RICH HUNDREDS OF MILLIONS MILLIONS evolving as we Network of Sensors of Mobile Devices of Mobile Users discover new technologies 17 Nokia Internal Use Only
  18. 18. …and Needs to Become Personal… Synthesis How A of Us as Who When Individuals Living Interactions Map Where with the Real World What 18 Nokia Internal Use Only
  19. 19. …to Everyone 19 Nokia Internal Use Only
  20. 20. Using Different Types of Data… Asserted Observed Reference Derived Data Data Data Data …to Create a Human Motion Graph 20 Nokia Internal Use Only

×