Telstra deploys Hadoop to pre-empt network issues

By on
Telstra deploys Hadoop to pre-empt network issues

Predictive analytics project to boost availability.

Telstra has deployed a predictive analytics system that uses its operational data to help identify network issues before they occur.

The big data facility is managed by CTO Vish Nandlall, who described it as “one of the largest Hadoop infrastructures in Australia”.

Telstra has not revealed the cost of the facility, which has been rolled out progressively during the past nine months.

“Conventional network problem solving generally only responds to events that have occurred,” a Telstra spokesperson said.

“The ambition of our big data program is to prevent incidents before they happen, and shorten response times when they do happen through automated response."

Telstra has created performance “baselines” based on the operational data it has stored.

It compares real-time network data against those baselines. Should network performance fall short, the analytics system triggers targeted warning messages and suggests potential responses.

Nandlall told iTnews the big data facility would also improve “capital deployment, transmission, infrastructure – how to better plan truck rolls, for example. How to better handle calls to our call centre.”

Telstra has been using predictive analytics to tackle customer churn since 2013, but that capability is managed by the marketing department, not IT.

The telco said its new big data project had been driven through the CTO team because “predictive and prescriptive analytics are technology-intensive capabilities, requiring scale-out compute and storage, data modelling, pattern matching, complex event processing, machine event and log indexing and search, and multidimensional analysis databases”.

Stay tuned for iTnews' CXO Challenge feature on how Telstra is approaching digital disruption.

Got a news tip for our journalists? Share it with us anonymously here.
Copyright © . All rights reserved.

Most Read Articles

Log In

  |  Forgot your password?