Telecommunication Sector: Improved business operations and planning efficiency by transforming call data record (CDR) data to real-time business intelligence system supporting decision making across the various business units of the company

The Problem
A telecom giant, with over 5 million subscribers has an older legacy business intelligence system which can no longer cope with handling the huge volume of input data files. Reports generation takes hours to produce, and often end up being inaccurate. Daily man-hour spent on report generation was enormous leaving the supporting IT staff with little or no time for other critical IT functions. Revenue assurance, marketing, product development etc. teams do not the need visibility to function properly.

Typically, about 10 data files per minute are extracted from the CDR files and sent for analysis. On the average, each file is over 10 Mbytes in size, and contains over 250,000 records. Hence, the solution is to process over 250 million call data records per day - from just one data-source alone. The files are to be processed as soon as they become available, so that dashboards and scheduled reports will be updated in near real-time.

This telecom giant turned to UnlimitedAnalytics to address this critical business problem.

The Solution
UnlimitedAnalytics is data-source and data-format independent. Hence, there was no need to introduce any new business logic; the system was only configured to map their data-sources (CDR data files - the main) for real-time uploading, ingestion, and reporting generation, visualization, and distribution (manual and automated):

• Which UnlimitedAnalytics solution implementations is required? Choosing between using the UnlimitedAnalytics cloud service or have it implemented on-premise: the company preferred on-premise implementation on their infrastructure (private cloud) - this was particularly preferred as the size of the call data record files generated every 10 minutes is huge.
• Identify all data-sources that were to serve as data input for the integrated enterprise data analytics and reporting solution.
• Identify fields within the CDR file record that are not required in any report, such were excluded from ingestion, and analysis during the data-source configuration phase
• Identify the levels of the data aggregation desired by the various business units
• Identify the internal and external reports needed by the various business units; when and how they need these reports to be composed, published and scheduled for automated generation and distribution.
• The CDR files is pushed to the UnlimitedAnalytics system every 10 minutes
• Data Ingestions commences once data files becomes available, the process is completed before new files are pushed
• Identify the various Dashboard views needed by different business units. These should be refreshed every 10 minutes to display new report based on the last data ingestion process
• Identify which fields in the CDR file is needed in any analysis
• Identify all the data sources that are required to generate the reports

The Results and Benefits
We provided a unified business intelligence system that provides visibility to business activities through the provision of on-demand and scheduled reports needed for operational, tactical, and strategic decision making at all levels of the business towards better service provisioning, increased profitability, and related benefits:

• Provided deeper insights into business and service performance for better, faster, and proactive decision making
• Eliminated manual effort to internal and external report generation
• Provided functional-units tailored real-time dashboard for service and business monitoring
• Provided accurate report that resulted in users confidence
• Business users can now author / compose, generate, publish, distribute, and schedule reports (graphical and tabular) without recourse to IT department as no coding or scripting is required, just click-and-select.
• Users can now compose reports based on their desired criteria, and aesthetics rather than being limited to standard predefined reports
• Some reports had their generation time reduced from days to minutes, and provided expected accuracy
• Composed, published, and scheduled the key business periodic (hourly, daily, weekly, monthly, quarterly and yearly) reports that were automatically generated based on dynamically specified date range
• Provided in-depth visibility to product offerings, subscribers, regions, locations, and base stations for service usage and profitability. Providing the executive summary report like: Top N, Bottom N, Between N1 and N2, > N, < N etc.
• Provided comparative and statistical analysis of chosen metrics, and providing the future forecast value for proactive planning
• CDR files were pushed and ingested every 5 minutes on 24X7, and the chosen aggregated methods of minute, hour, day, week, month, quarter, and year levels were updated immediately; and becomes available in all reports and dashboards
• Provided system that is open for users to add new data sources without recourse to us
• Provided access to data-sources, and reports based on need-to-know basis specified by report author
• Provided comprehensive logs and audit trails for system, and users activities