Presented By O'Reilly and Cloudera
December 5-6, 2016: Training
December 6–8, 2016: Tutorials & Conference

From telco data to real-world data analytics products at SmartHub

Boon Siew Seah (SmartHub, StarHub Ltd.)
11:15am–11:55am Thursday, December 8, 2016
Location: 308/309 Level: Beginner
Average rating: ****.
(4.00, 2 ratings)

Prerequisite Knowledge

  • General knowledge of real-time and batch processing big data technologies

What you'll learn

  • Understand the value of telco analytics, what works and what doesn't when building viable telco analytics products using real-time telco data, and how SmartHub built its big data stack to meet customers' SLA requirements


Translating streaming, real-time telecommunications data into actionable analytics products remains a challenging task. From a technical standpoint, the volume of generated data, the complexity of models, and the need to deliver insights at near real time demands a carefully designed big data architecture. Beyond the technical requirements, you must also address the challenge of delivering actionable insights to customers while protecting the privacy of personal data.

SmartHub, the analytics division in StarHub, has built and delivered telco analytics products to its customers across various industry verticals that range from the Singapore government to retail mall owners. Boon Siew Seah explores SmartHub’s past successes and failures building telco analytics products for its customers and shares the big data technologies behind its two API-based telco analytics products: Grid360 (geolocation analytics) and C360 (consumer insights). Grid360 delivers streaming geolocation insights derived through StarHub’s streaming and nonstreaming telecommunications data; C360 builds consumer demographic, psychometric, and behavioral profiles based on telecommunications data insights.

Topics include:

  • An introduction to Grid360 and C360, with case studies on the value they create for customers
  • An overview the analytic models behind Grid360 and C360 and the technical challenges encountered while building them
  • The analytics architecture, which is designed to meet customers’ service level agreement (SLA) requirements (latency, fault tolerance, persistence, and scalability)—with a big data stack comprised of Apache Hadoop, Spark Streaming, Kafka, and Cassandra
  • How and why SmartHub settled on the above technology stack
  • Past successes and failures on converting telecommunications data into viable data analytics products
Photo of Boon Siew Seah

Boon Siew Seah

SmartHub, StarHub Ltd.

Boon Siew Seah is the head of data engineering and innovation at SmartHub (part of StarHub Ltd.), where he is responsible for driving data innovation through telco data and big data technologies to deliver unique telco analytics products. Prior to joining StarHub, Boon Siew was the data science guild lead in the Data Science division of Infocomm Development Authority (IDA) in Singapore.