Artificial intelligence is a beacon of progress, bringing data integrity, scalability and security to the forefront in today's dynamic public sector landscape. The ability to make data-driven decisions and deliver efficient services is critical, but protecting citizen privacy remains a top priority.
In a new interview with FedScoop, Dr. Kimberly McManus, deputy chief technology officer for AI at the Department of Veterans Affairs, and Andrew Davidson, SVP of products at MongoDB, discuss how combining AI with modern tools can help public agencies better serve their communities while maintaining trust.
For example, McManus shared how the Veterans Health Administration (VHA) is actively exploring AI technologies to streamline workflows and improve efficiency. The agency is currently working on a number of pilot projects, including one to automate the process of recording clinical interactions by recording doctor-patient conversations, transcribing them, and summarizing them into notes that doctors can review before adding them to patients' electronic health records.
The other is community care document processing. Currently, VHA receives a lot of documents from external sources that need to be integrated into the medical record. VHA wants to use optical character recognition and extraction technology to streamline the processing of these documents.
“Both of these projects are still in their early stages but are showing promising results and significant increases in staff time savings,” McManus said.
In addition to these efforts, public agencies should also consider how to integrate AI to enhance data management. McManus emphasized the importance of selecting the right data management platform to support different use cases, such as a transactional database for operational processes and an enterprise data warehouse for AI and data analytics. He also emphasized the importance of a robust data catalog to identify available data types and exploring new database technologies, such as vector databases for AI pilots.
Meanwhile, Davidson discussed the evolution of data management platforms in the AI era, emphasizing the importance of data modeling and vector search capabilities, and integrating these into software applications to maximize the benefits of AI. “Not only are data solutions themselves evolving, but team structures are evolving too,” Davidson said.
McManus and Davidson also discussed the importance of data privacy. For example, VHA is working on a number of fronts to ensure privacy-preserving data analytics, including differential privacy, removal of personally identifiable information and protected health information, synthetic data, and federated learning. These methods aim to balance data accessibility with strict privacy requirements.
Davidson added that modern data solutions such as variable encryption and role-based access control are essential to maintaining data privacy in AI applications. He also noted the benefits of leveraging cloud-based services, which offer a higher level of abstraction and allow smaller teams to work more quickly while maintaining security.
Learn more about how MongoDB can accelerate and simplify your mission here. It serves the entire public sector, including federal, state and local governments and defense agencies.
This video panel discussion was produced for FedScoop by Scoop News Group and sponsored by MongoDB and AWS.
