What is Analytics Modernization?

Analytics modernization is an effort by Pitt IT to continuously improve the way the university uses data for decision-making and operations. The effort is split into three focus areas.

Modernizing Our Data Infrastructure

Pitt IT Analytics is working to create a new home for the University Data Warehouse (UDW) in AWS using data lake technology. The data lake technology allows for storing all kinds of data - traditional structured data, as well semi-structured data or objects like documents, images, or video. This flexibility will allow for more kinds of data to be made available, as well as laying the foundations for machine learning and artificial intelligence (AI/ML).

Migrating to the AWS cloud also provides other advantages. AWS features numerous off-the-shelf tools that can easily connect to Pitt’s data stored there. Tools like Sagemaker allow data scientists to build, train, and run models. Other services, like Comprehend, can work with text to identify sentiment, keywords, or topics. AWS adds new tools all the time, increasing our ability to do more with our data.

The data marts from the current Oracle UDW will be migrated to the new AWS platform, but will still retain the same structures, so users will be able to easily adjust any queries or reports running against the Oracle data.

Migrating Workloads to Tableau

Currently, Pitt IT analytics supports three tools for business intelligence, data, and analytics - Tableau, Oracle Analytics Cloud (OAC), and Cognos. Tableau has proven to be the most popular of the tools, and has proven that it can work with diverse university data - from general ledger to student data.

Pitt IT Analytics will be making available new Tableau products - data sources, starter dashboards, and enterprise dashboards across data domains. These new offerings will over time replace existing reports and dashboards from Cognos and OAC. As users migrate their workloads to Tableau, parallel data and reporting from Cognos and OAC will be decommissioned, leading to reduced costs and the potential to eventually stop support of those tools.

As part of the migration to Tableau, Pitt IT will provide training, support, and documentation to help users learn Tableau, and empower end users to do more with their data.

Empowering Our People

Our data and tools are only part of the equation. Our most important resource is our people, and as part of the Analytics Modernization effort, we’ll be rolling out training and programs to empower Pitt’s data users. As mentioned above, training our people on using Tableau is a critical part of improving our data usage. But another very important aspect of empowering our people is recognizing that not all data users are the same.

The Data Ranger program will bring together data analysts, data scientists, and data-intensive technologists to push beyond operational and data silos and work toward common practices, tools, and documentation. This group will act as a sounding board for data governance efforts, and advocate across campus for data-informed decision making and data literacy.