Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Analyze Telecom Fraud at Hadoop Scale

835 views

Published on

Analyze Telecom Fraud at Hadoop Scale

Published in: Technology
  • Be the first to comment

Analyze Telecom Fraud at Hadoop Scale

  1. 1. Page1 Diyotta, Inc. All Rights Reserved Analyze Telecom Fraud at Hadoop Scale 29th June 2016 Sanjay Vyas Co-founder & COO, Diyotta
  2. 2. Page2 Diyotta, Inc. All Rights Reserved Telecom-Relevant Glossary • CDR – Call Detail Record • Any phone call generates a CDR • IPDR – IP Detail Record • Any internet browsing activity generates an IPDR • IVR – Interactive Voice Response • Automated telephone response system usually for typical queries
  3. 3. Page3 Diyotta, Inc. All Rights Reserved Fraud in Telecom Global Mobile Industry $2.2T Revenue Losses due to Fraud $46.3B Reference: http://www.argyledata.com/files/Telecom-Fraud-101-eBook.pdf
  4. 4. Page4 Diyotta, Inc. All Rights Reserved A Day In Life of Telecom Data (Fraud Use-Case) Source Systems Ingestion Pipelines Target Data Sets Fraud Analysis Minataur
  5. 5. Page5 Diyotta, Inc. All Rights Reserved Legacy State for Fraud Analytics Monolithic Script Based file Ingestion Minataur Fraud Application • Limited capacity for processing • Cannot Scale for Volume/Velocity • Cannot do on-demand real-time Analytics
  6. 6. Page6 Diyotta, Inc. All Rights Reserved Business Challenges • Business not able to • analyze IPDR Data due to the sheer volume • ingest streaming data from IVR systems for fraud analysis • Perform on-demand real-time fraud analytics for deeper insights
  7. 7. Page7 Diyotta, Inc. All Rights Reserved IT Challenges with Hadoop Adoption • Skill Gap • Limited in-house expertise on evolving technologies and keep up the pace • Enterprise Standards • Manual coding suffers from quality/maintenance issues and is inconsistent • Scalability across data and technology • Real-time, social media, multi-processing engines • Data Lineage
  8. 8. Page8 Diyotta, Inc. All Rights Reserved Solution Components
  9. 9. Page9 Diyotta, Inc. All Rights Reserved Solution Architecture for Fraud Use-Case
  10. 10. Page10 Diyotta, Inc. All Rights Reserved Diyotta Modern Data Integration Platform
  11. 11. Page11 Diyotta, Inc. All Rights Reserved Page11 Diyotta, Inc. All Rights Reserved Diyotta Architecture
  12. 12. Page12 Diyotta, Inc. All Rights Reserved Page12 Diyotta, Inc. All Rights Reserved Customer Success Story
  13. 13. Page13 Diyotta, Inc. All Rights Reserved Q&A Sanjay Vyas Email: sanjay@diyotta.com Web: http://www.diyotta.com Trial: www.Diyotta.com/try

×