Case Study

Summary

Since 2020, Nava has partnered with the Commonwealth of Massachusetts to help build and maintain the state’s paid family and medical leave (PFML) program. In just 13 months, we helped Massachusetts design, build, test, and launch this new digital service that helps workers apply for and manage PFML benefits. As part of this work, Massachusetts had to integrate custom technology — like the applicant and employer website that we helped build — with commercial products, including an application processing system. We helped Massachusetts build the Paid Leave application programming interface (API) and a back-end API layer that enables several disparate state systems to communicate with the PFML website and with one another. As a result, Massachusetts’ PFML program is more flexible, resilient, and human-centered. 

Approach

When building the API layer, Massachusetts wanted to minimize risk while delivering an excellent service and meeting an aggressive launch deadline. To work smarter and faster, we partnered with Massachusetts and other vendors to iterate on the API layer through a series of pilot releases. This helped us ensure it would meet peoples’ needs when launched. We also planned a soft launch several weeks before the legislative deadline to guarantee everything worked before scaling the service. 

The Paid Leave API is one component of Massachusetts’ PFML API layer — or a larger system of APIs, a database, technological infrastructure in Amazon Web Services (AWS), and a framework that enables batch data processing. Of the whole PFML API layer, Nava’s role included building many of the batch jobs and their framework, the database, and supporting infrastructure in AWS. Other vendors led end-to-end testing of the API layer, also built batch jobs, and currently maintain and improve the database and infrastructure in AWS.

Outcomes

The API layer enables easy integration between other IT systems that support the PFML program and enables Massachusetts to easily adapt new tools and workflows based on the changing needs of the agency and its constituents. This ensures a more accessible and equitable service. 

Below are three examples of how the API layer has helped improve outcomes for agency staff and applicants. 

Reducing administrative burden 

Staff at Massachusetts’ public-facing call center previously had to go into the application processing system to authenticate — or verify the identities of — applicants that called with questions. This was an extra manual step that slowed down the customer service experience for applicants. An improvement that Nava helped Massachusetts make to the API layer addresses this by connecting the application processing system to the call center system, enabling staff to authenticate callers faster and more efficiently. On the day that Massachusetts released this change, they received over 4,000 calls — a record number at the time. The API layer improvement led to a reduction in average call time.

The Paid Leave API also enables agency staff to securely receive accurate, real-time applicant data that can help them adjudicate applications. This reduces the administrative burden of manually transferring data and promotes timely payment of benefits. 

Empowering agency staff with data 

The API layer transmits data to agency staff.  These resources can reveal insights on everything from technical issues to the efficacy of new features to how systems handle uncommon cases. Agency staff can then use these insights to iterate and continuously improve on the program.

Continuously improving as the program matures

We have supported Massachusetts’ efforts to continuously improve program operations, customer service, and graceful scaling.

For example, the API Layer has helped refine business processes over time by enabling people who begin a PFML application over the phone to manage it on the website, too. The API layer also lets people make changes to their PFML application and file appeals online, rather than needing to contact the call center. Both of these improvements have helped serve more people faster.  

Finally, the API layer allowed timely implementation of Massachusetts’ “top-off” legislation, a new policy that allows employers to supplement PFML benefits with other leave like paid time off (PTO). 

Process

To help build the PFML API layer, we began by prioritizing fundamental integrations that would have to work in order for the rest of the program to succeed. Examples of prioritized integrations include Massachusetts’ PFML application processing system and non-PFML government databases essential to adjudicating applications. 

We also took a human-centered approach and partnered with stakeholders to understand applicant needs and program priorities. This alignment helped us continuously iterate on the parts of the API layer we helped build and de-risk delivery. It also helped us understand Massachusetts’ operational needs, which we designed the API layer to serve long-term. 

During the early phases of development, we worked with a third-party vendor to test the API layer several times in a non-production environment, meaning our tests wouldn’t interfere with live applicant data. These tests helped us identify and address risks early on and iterate based on our findings. 

Next, we moved onto our user-facing pilot, where real applicants seeking medical leave were able to create an account on the website, login, and submit an application. We used these applications to perform tests of key integrations and to better understand how the Paid Leave API would have to function for different applicant scenarios. After honing the API layer over a series of pilots, Massachusetts released it to the public alongside the application website. 

Conclusion

Through user research and iteration, Massachusetts was able to build an API layer that makes their PFML program more resilient, secure, adaptable, and human-centered. The API layer helps connect the applicant and employer website to various external systems, reducing administrative burdens on agency staff, promoting timely payment of benefits, and enabling agency staff to leverage data for their internal operations. We continue to help Massachusetts iterate on the API layer to meet their constituents’ changing needs, fostering increased trust and a more equitable service.