How To Implement A Successful Data Strategy: A Success Story
Technological advancements in the past several years have accelerated dreams coming true for many in the data world. The ability to store huge amounts of data in clouds and access it all very quickly has been a game-changer. We’re not limited to leveraging data that is days, weeks, and sometimes months old to make decisions. These dreams come true have also created some nightmares. We now have more data than we know what to do with. As more data is available in real-time, we’re asking more and tougher questions. Demands on data analysts have become unrealistic. We’re overwhelmed by possibility and a fear of missing out on the next big insight. This saga is familiar for many in the data world, and USCIS is no exception. While the challenge is complex, the solution is amazingly simple. Focus on your mission. Why does your organization exist? And how can data solve the challenges that prevent your organization from operating better.
When I joined USCIS in 2018, I was blessed to join an organization that was naturally data-driven; USCIS knows how valuable our data are and are not afraid to invest in getting the most out of it. Consequently, I was blessed to find an organization that had already migrated all of its data into the cloud! At that time, such a concept was practically unheard of in the federal government.After launching the first USCIS Data Strategy in 2019--which doggedly focuses on the USCIS mission—we had a rally cry around four goals aimed at making USCIS data reliable, comprehensive, and used to make informed decisions at strategic operational levels. After three years of intensive investment, USCIS has made huge strides in accomplishing those goals as we near the launch of our second USCIS Data Strategy.
Technology has enabled advancements in all four goals:
1. Data Management: Modernize data infrastructure and improve data management practices–Although USCIS was already in the cloud, we needed to meet growing demand for data with a dynamic and scalable platform. Apache Spark with Databricks, cloud native data lake with Delta Lake, MLflow, and other tools became a crucial factor for the success of our programs.Deploying a new Data Migration Service (DMS) has enabled more frequent data refresh into our data lake. Now, all of our data is accessible from one place! With so much data, governance and knowing what we have can be a challenge. Master data management tools like Collibra, which automatically update data standards, policy, and access controls, allow us to safely leverage all that data.
2. Business Intelligence: Advance business Intelligence at all levels of USCIS–USCIS is no stranger to statistical reporting. We have extremely skilled statisticians and data analysts that query data stores thousands of times a years. Data visualization tools like Tableau have transformed static spreadsheet-centric reports into dynamic, interactive windows of insight. Connecting Tableau to our enterprise data lake, which enables automatic updates, has increased the accessibility of data immeasurably. I firmly believe that it is easier to draw insights from a graph than it is from a table of thousands of data points. I also believe that people are more likely to seek out data that they know is fresh. That hypothesis has born out as the Director’s Catalog and its 18 dashboards have more than 100,000 views in the past six months. These tools make citizen-development very easy to manage and federate the governance required to maintain so much knowledge. These data visualization tools have also offered opportunities to monitor data quality in real time and have become integral components to our ability to ensure that our data meets high integrity standards.
"While technology has provided the infrastructure to realize these changes, it is really our people that have embraced the technology to change the analytical landscape of USCIS"
3. Network Analytics: Strengthen adjudication through discovery of relationship networks– Historically, USCIS case management systems have designed to record transactions and lack a person-centric view on applicants. Resolving millions of immigration records is no small feat!Taking advantage of the tools in Databricks, Graph Frames and Graph x, and in Tableau, makes it shockingly easy to visualize the obvious and non-obvious relationships among the different actors in the immigration system. USCIS has tested this capability on several use cases and looks forward to expanding its utility.
4. Optimization: Optimize operator/ analyst time by preprocessing data with assistive technologies– Although it has taken a long time to get the security right around robotic process automation (RPA) tools like UiPath, RPA have saved thousands of man hours by automating wrote manual tasks like XYZ. It has also served to provide fast, short-term solutions that bridge between now and the future where we have built new, sustained capability in our source systems.
While technology has provided the infrastructure to realize these changes, it is really our people that have embraced the technology to change the analytical landscape of USCIS. Our analysts and staff have adopted the products that are enabled through technology and helped the data to shine. USCIS has a healthy cadre of data analysts who are hungry for more knowledge and capability. Our staff have embraced the analytical products that are produced to improve their decision-making and influence their behavior. Our data literacy program has offered resources and a training curriculum that provides a consistent baseline of critical thinking and analytical skills for our analysts AND a network of other analysts upon whom they can call for help. The data literacy program is also expanding to reach every corner of USCIS because, these days, data is part of everyone’s job.