Case Study

Samsung

Download PDF     See a Demo
Company Profile
Company:
Samsung
Size:
Large Enterprise
Industry:
Manufacturing
Location:
Seoul, South Korea

Background:

Samsung has been analyzing and improving mobile and smart TV products through product analytics for decades. Through that period, a number of different tools, approaches, data repositories, data capture, and data storage locations have been employed. To improve product performance, reliability, feature adoption, ease of use, Samsung has captured realms of device-specific data, such as the location, hardware specifications, utilization rates, capacity and battery life, etc. Hadoop makes the processing, collection, analytics of this data faster and move cost effective for Samsung.

Challenges:

As a global manufacturer with products in all markets and territories, Samsung also must adequately protect any sensitive data from device logs and data capture. Specifically, in Europe, new privacy policies defined in the European Union Privacy Directive require Samsung to protect any personal identifiable information specific to European citizens. Samsung still needed to collect device data for analytics, but was mindful of privacy laws, and privacy fines levied on competitors that did not fully comply with privacy mandates.

Use Cases:

  • Aggregate logging data (product, usage, user configuration) for all smartphones worldwide
  • De-identify personal user info to ensure privacy and compliance with European/US Privacy
  • Keep all sensitive data encrypted at-rest, and provide authorized access (decryption) of sensitive data on a case-by-case basis for analytics applications that require access to full, complete, plaintext data.

Results:

  • Samsung utilizes PKWARE Flume agent to protect all sensitive data being written to Amazon S3
  • Runs PKWARE in AWS, also utilizes PKWARE EMR security agents to selectively decrypt for authorized analytics in AWS
  • Achieves On-demand Hadoop for product analytics, user behavior, supply chain optimization in high scale-out, high performance and high availability system
  • 100% cloud based