Tag Archives: BIGData

The Big Data Strategy



For data analytics, T-Systems differentiates by volume: small (GB), big (TB), huge (PB).

Due to the multi-platform strategy pursued by T-Systems, we can offer solutions in all the three volume categories, ranging from private, public onto dedicated, dynamic cloud platforms.

The primary platforms/products offered by T-Systems are: Cloudera Hadoop, Talend, Tableau and Microsoft Azure.

The broad strategy is outlined below.



T-Systems is vendor-neutral, and strives to offer the most appropriate solution for its customers.

We start with analyzing the customer situation, making a gap analysis for the future scenario, analyzing a couple of possible solutions, and recommending the best fit.

Customer Story

Let us assume that the customer comes from the Manufacturing segment, and wants to analyze machine data in a factory premise. First, we prepare a customer story, that describes the customer situation, the goal, and the means to achieve it.

A Manufacturing Customer Story could look like this:


For demo purposes, we often use Microsoft Azure with PowerBI for dashboard features.

Sample dashboards for Manufacturing customer are shown below, prepared in PowerBI, with data analysis in Microsoft Azure.







The principal process chain behind our approach is as follows:


Ingest: this phase collects the data from the physical world into the analytics platform. The physical world comprises field sensors and machine parts, but also social media like Twitter updates. This process also comes out to be the most complex, as integration of varied kinds of sensors adds complexity to the system.

Prepare: this phase cleanses the data, identifies gaps, gives structure, and prepares it for storage in the data platform.

Store: this phase actually stores the data. Two primary modes are supported: hot (real-time) and cold (for deeper, more complex analytics in a longer time period).

Analyze: this phase is the core analytics engine, where much of the data processing takes place. Insights from data are generated in this phase. This is normally the most expensive component in the analytics chain.

Output: this phase is the customer interface, thus the most visible part of the analytics chain. Typical output modes are computer screen and printed reports, but of late, tablets and mobile phone notification have gained immense importance.



In a typical T-Systems offering, there are 3 price components, as shown below.


Product Architecture is a one-time activity, at the beginning of the project, and is charged as a fixed one-time fee.

The System Implementation is charged by man-days, and can be compared to a consulting offering.

The Managed Hosting is a running cost, charged per month, for a managed cloud environement, hosted and managed end-to-end by T-Systems for the customer.