How AI Helped a Federal Health Agency Transform Medical Device Reporting

The digital healthcare landscape has been rapidly changing in North America after the COVID-19 pandemic due to sweeping demographic, policy, and technological changes. With these changes in mind, timely access to accurate data can mean the difference between proactive intervention and delayed response for patients. At Improving, we’ve helped numerous healthcare organizations, departments, and agencies navigate the complexities of the digital healthcare landscape through our expertise in AI-powered solutions. One such example involves a federal health agency in North America that faced a mounting challenge: manually processing thousands of medical device adverse event reports submitted by manufacturers. These reports, often non-standardized and emailed in bulk, were manually triaged and entered into internal systems – a very a labor-intensive process that created backlogs, delayed risk detection, and increased the potential for human error. 

Our Google Practice team saw an incredible opportunity to bring the power of artificial intelligence and cloud computing to bear on this critical public health workflow. The subsequent collaboration with the agency resulted in a transformative solution that not only streamlined operations but also enhanced its ability to protect citizens from device-related health risks. 

The Problem: Manual Processes and Mounting Risk 

Medical device manufacturers are required to report adverse events to our client. These reports arrive in varied formats and are manually processed by staff; the lack of standardization and automation led to significant inefficiencies. For example, reports were emailed to a central inbox, where they were manually reviewed, triaged, and entered into internal systems. This process was inherently slow, error-prone, and unsustainable given the growing volume of submissions. 

The consequences were serious: delays in identifying potential health risks, increased operational costs, and a growing backlog that threatened regulatory compliance and public safety. 

Our Approach: AI-Powered Automation 

In response to our client’s challenges, Improving developed an Automated Medical Device Reporting Solution that leveraged Google Cloud technologies and machine learning to automate the entire data ingestion and processing workflow. The system included: 

• A front-end web application for form submission and review 

• Back-end systems for data processing and storage 

• Machine learning models for intelligent data extraction 

• Document AI for parsing non-standardized forms 

• BigQuery for scalable data analytics 

• API integration for seamless data transfer to internal systems 

This architecture enabled rapid, accurate processing of diverse form types, transforming a weeks-long workflow into one that could be completed in hours. 

Business Benefits: Efficiency, Accuracy, and Insight 

The impact of the solution was immediate and profound: 

• Enhanced Efficiency: Processing time dropped from weeks to hours, clearing backlogs and freeing up staff for higher-value tasks. 

• Improved Accuracy: AI models reduced human error, ensuring reliable data for regulatory decision-making. 

• Scalability: The system can handle increasing volumes without additional manual labor. 

• Cost Savings: Automation reduced the need for temporary data entry personnel, delivering long-term savings. 

• Regulatory Compliance: The solution met all data protection requirements, including handling of sensitive information. 

• Data-Driven Insights: Structured data enabled faster identification of patterns and potential device issues. 

Technologies and Methodologies 

Our solution was built on a foundation of cutting-edge technologies on Google Cloud and agile development practices that include: 

• Google Cloud Platform: Provided a secure, scalable infrastructure 

• Document AI: Enabled intelligent extraction from diverse form formats 

• BigQuery: Delivered fast, reliable data processing 

• Machine Learning: Powered high-accuracy classification and validation 

• API Integration: Ensured seamless interoperability with internal systems 

• Agile Methodology: Supported iterative development and rapid adaptation to feedback 

Partnerships That Made It Possible 

A high degree of collaboration was key to success on this project to ensure that our team understood our client’s needs and delivered a solution that fully addressed their challenges in an efficient, compliant, accurate, and scalable manner. We worked closely with: 

• Google Cloud’s public sector engineering team and the Document AI product team, whose expertise helped overcome implementation challenges 

• The agency’s corporate IT and security teams, ensuring compliance with privacy and security regulations 

• Procurement partners, who facilitated contracting and ensured all personnel met government security clearance requirements 

Lessons Learned: Insights for Public Sector AI Projects 

Our team had the opportunity to learn many valuable lessons for future healthcare and AI-related projects, which were: 

• Data Centrality: Early access to representative datasets is critical for training and validation 

• Scope Management: Changes in data formats can impact timelines—clear scope definition is essential 

• Client Collaboration: Continuous communication accelerates issue resolution and ensures alignment 

• Iterative Development: Agile methods foster adaptability and responsiveness 

• Government Navigation: Patience and documentation are vital for navigating public sector processes 

• Compliance Rigor: Thorough documentation and adherence to regulatory standards are non-negotiable 

Conclusion: A Blueprint for AI in Healthcare 

Improving’s work with our federal health agency clients demonstrates how AI can revolutionize public sector healthcare operations by leveraging leading-edge technologies on Google Cloud. By automating a critical workflow, we helped the agency respond significantly faster to potential health risks, improve data quality, and reduce operational costs. This project is a testament to our ability to deliver impactful, compliant, and scalable solutions that serve the public good. 

As governments and healthcare agencies continue to grapple with data overload and operational inefficiencies, this case study offers a compelling blueprint for transformation. We’re proud to be at the forefront of AI innovation in healthcare—and ready to help more organizations unlock its potential.