Making data matter at Mathematica

Mathematica’s Akira Bell is a CIO with a mission.

One of three finalists for the prestigious 2024 MIT CIO Leadership Award, Bell led the development of a proprietary data and analytics platform on AWS that enables the company to serve critical data to Medicare and other state and federal agencies as well as the Bill and Melinda Gates Foundation.

Formerly known as Mathematica Policy Research, the Princeton, N.J.-based research firm is proud of its mission to deliver accurate data to ensure goods and services are distributed with equity and precision in a socially just manner. It is one of the few midsize companies with Federal Risk and Authorization Management Program (FedRAMP) authorization, the government’s highest security certification for cloud operators and required for work with federal agencies.

“Akira exemplifies the traits we want to see in a strong IT leader,” says George Westerman, senior lecturer at MIT Sloan School of Management, founder of the Global Opportunity Initiative, and a judge for this year’s MIT CIO Award. “She and her team are making big changes happen in an organization focused on providing analytic evidence-based insights for public policy questions.”

Bell was honored to be among the other nominees from large enterprises Home Depot and ServiceNow and attributes her finalist standing to her 70-strong IT team, the company’s transformation to cloud computing, and innovation of its Mquiry analytics platform.

“We are one of very few companies with a platform-as-a-service that is now FedRAMP-compliant and not a major technology company,” Bell says. “That was not a small effort.”

Emphasizing ethics and impact

Like many of the government agencies it serves, Mathematica started its cloud journey on AWS shortly after Bell arrived six years ago and built the Mquiry data collection, collaboration, management, and analytics platform on the Mathematica Cloud Support System for its myriad clients.

Bell oversees two CTOs, data scientists, researchers, social scientists, and cloud and security experts with an aim to improve the efficiencies and value of services Mathematica delivers to less advantaged members of society. 

“We set the vision together,” Bell says. “I’m reliant on two CTOs to bring the external signals in, reliant on our product team to learn what advanced technologies we need to bring in to meet some of those external signals, and on a cloud and security team to keep us heavily centered on digital trust.”

The company’s three business units — healthcare, human services, and global business —provide critical insights to national health agencies such as Medicare and Medicaid, as well as rural healthcare systems. In addition, it works with education, nutrition, and child services data in serving agencies such as the Administration of Children and Families, as well as climate data when working with agricultural organizations.

Mathematica employs roughly 500 social scientists and researchers and about 130 data scientists, Bell says.

“A lot of our work is for agencies who are trying to help people who don’t have a lot of wealth,” Bell says. “We think about ethics and governance and making sure we are not amplifying any bias. Our clients are asking us to solve problems like access to health care, maternal health care, or child welfare and making sure there’s equity in the outcomes.”

The power of partnership

Bell’s predecessor built a strong networking and security infrastructure, while her role has been focused on migrating to the cloud, building the analytical platform, and developing new client-facing professional services. Like others, Bell’s data scientists face challenges such as data cleanliness and interoperability, and Mathematica will at times partner with other organizations to overcome those challenges.

For example, Mathematica partnered with New Wave to develop a cloud-based data quality platform, Imersis, which helps state Medicaid agencies meet required data quality and certification requirements. The tool has been adopted by New Jersey, West Virginia, and the US Virgin Islands to help address quality issues and improve data submission to the Centers for Medicare and Medicaid Services (CMS) and Children’s Health Insurance Program (CHIP) Services.

Errol Blake, director of product innovation and strategy at New Wave, says both companies share the same core values of integrity and social responsibility and developed the tool to help state Medicaid agencies meet Transformed Medicaid Statistical Information System (T-MSIS) data outcome-based assessments and Medicaid enterprise system certifications requirements.

“The Centers for Medicare and Medicaid Services [have] been talking about T-MSIS data quality in the states for 20 years, and we saw the opportunity to send good quality data to CMS, which wants to ensure that every state is operating efficiently, effectively, and the beneficiaries of Medicaid are getting the service that they truly deserve,” Blake says. “It was a marriage made in heaven.”

As for what the partners bring to that marriage, Blake says: “Mathematica brings the knowledge of Medicaid policy and data, and New Wave [brings] technical innovation and our HITRUST certification, which is the highest security certification one can achieve.”

AI ‘bake-offs’ under way

Mathematica’s PaaS has not yet implemented AI models in production, but Bell grasps the power of machine learning (ML) and generative AI to uncover new insights that will help Mathematica’s clients. 

Mathematica’s AI Task Force, Bell says, is evaluating many ML and gen AI technologies in a “bake-off” behind closed doors, but data governance, data security and ethical considerations will be paramount before integrating any AI into its research and analytics.  

“We’re still in experimentation mode,” Bell says, noting she is investigating LLM training models for internal use. “Once we get more data from across a couple of areas into Mquiry, I would love to see the insights it might show us and do some training against that data. We’ve been working towards both and I think there are bigger opportunities around global where we have a lot of different types of data such as social determinants of healthcare, which could be a big use case.”