Streamlining Large Scale ETL in Telecom

The Problem

Our large telecom client was working through a multi-year project transitioning millions of customers across to another system. Despite having a plan in place to move data across, employees were stuck waking up in the middle of the night to monitor processes, check for new files, run complex custom queries, transform files, then distribute results several weeks each month during the process. All of the tasks that could be easily automated already had been, but what was left involved spanning multiple backend systems, applying complicated business rules that frequently changed, and according to published scheduled that often changed.

The Solution

The Eigen X team built a solution to automate remaining processes, monitor files according to established patterns during peak processing times, move dozens of gigabytes of data across during an evening, all fully automated. The team authored C# code to automate legacy mainframe green screen processes, built a custom C++ ODBC driver to pull data from a legacy system, and wrote a Java backend to manage schedules, queue jobs, and monitor directories and sites. Custom PowerShell and Perl scripts validated incoming and outgoing files against established business rules. Tools were supplied to allow the business users to control the overall process in terms of scheduling, email notifications, and specific file types.

The Result

The custom ETL solution creatd by the Eigen X team saved hundreds of hours of labor for key employees during overnight batch processing, improved accuracy by minimizing the chance for human keystroke errors, minimized expensive downtime in systems used by hundreds of people, and improved data quality by placing additional safeguards in place to catch potential errors before they were distrubted outside of the organization.

Case Study - Large University – Marketing and Student Data Management within Salesforce.com

Problem

A large university was challenged with student data in multiple systems, duplicate data, and data quality.  This prevented effective measurement of the Return on Investment of marketing campaigns. They were using multiple Microsoft applications to manage prospective and registered students.  These various applications were not integrated with the student information system nor any of the other marketing systems.

Each campaign required a nearly a week of manual data cleansing to identify targeted prospects. Due to combinations of list data and data quality issues, duplicate messages were frequently sent to the same students for a given campaign.

Solution

The Eigen X team worked closely with the client to develop a strategic data roadmap. We provided the current state of student data management map as well as a future state using Salesforce as the primary source for marketing all data. This solution leveraged out of the box functionality within the Salesforce campaign module.  This allows the university to easily generate targeted lists, improve data quality, eliminate duplicates, track results and generate ROI reports to monitor campaign effectiveness.

Marketing automation included the integration of Salesforce with Constant Contact, web forms from student registration portal and Google Analytics.

Results

With this solution, the university will eliminate at least four days of effort currently required to create accurate marketing campaigns. They are now able to generate high quality distribution lists and initiate a campaign in a single day. The results of each marketing effort can be continuously monitored to measure key ROI metrics and overall effectiveness of the campaign at level not previously possible.