Payroll Data as a Product

Understanding payroll data in its current form, preparing it for migration to a modern system, and liberating elements for expanded future use in reporting and analytics is foundational to payroll system modernization. Data as a Product is the minimum data necessary to run payroll that is trusted, accessible, and reusable. 

PDaaP work includes things like: creating standard definitions for field use; determining data clean up needs; prioritizing future integration needs

The Payroll Implementation team has been doing 'data pre-work' for many months and in parallel with the selection of the vendor and new payroll system. This work expands on previous workshops to delve deeper into the people, processes and systems related to payroll. Agency partner and other stakeholder feedback helps shape this valuable work.

The COMPAS team has engaged with a variety of vendor partners to collaborate in the development and delivery of the modernized payroll system. 

• OIT: Several OIT Data Architects and members of the Colorado Digital Services team have been assigned to the Payroll Implementation Team. These are dedicated State resources working on data sourcing and profiling using Informatica as a tool in the validation of current system data. This team also is dissecting current system interfaces and integrations to ensure a seamless transfer to the new system. Find more information about OIT here.

• AdHoc: AdHoc was contracted to the Payroll Implementation Team in early 2023 and they are supporting the Payroll Implementation Team with data profiling, creation of data journey maps and user personas, creation of a data catalog and glossary, interface and integration research, and supporting data liberation work with the creation of a data lakehouse for future State analytic use. Find more information about AdHoc here.

• CGI: CGI is a vendor partner that currently works with and supports CORE and financial functions and processes throughout the state. Find more information about CGI here.

The current payroll data is stored in a complicated mainframe environment behind complex coding and formatting. With keen focus, the Payroll Implementation Team has taken a four-tiered approach in regard to the State's payroll data to understand integrations and interfaces.  These include: liberate the data in a manageable and secure format; evaluate the data using tools like Informatica, AWS RedShift, and Tableau; and, prepare users and the system for systematic and standardized data entry for a seamless transition to a new, modern system. 

Collaboration between agencies and data team

While the data team is preparing the data from CPPS to be ingested into the new COMPAS system, there are certain questions that have surfaced about the format and logic of the data. Since the data is coming from a system that is multiple decades old, there are likely to be a variety of anomalies that are not immediately understood by those interacting with the data.

To help answer these data questions, the data team has taken the opportunity to collaborate with Trail Guides, who are a subset of HR and payroll SMEs across a variety of agencies. Both teams meet on a regular cadence where the data team describes their questions about the data and the Trail Guides provide context and use cases to resolve the concerns about the data.

Below are some examples of artifacts the data team created to help describe the specific data concerns and questions that have come up.

Job Status Value “T”

Duplicate Positions

Reconciling Gross Pay, Net Pay, and Deductions

Employees with Multiple Original Start Dates

After the discussion, the data team will clean up the data in a way that reflects the information they learned from the Trail Guides.

This particular engagement is an important step in validating the data coming out of CPPS and promotes the opportunity for cleaner data to enter into the new COMPAS system.

The COMPAS team came together with vendor partners to: