Every year, workers’ compensation programs help hundreds of thousands of people (link resides outside of ibm.com) across the United States get back on their feet after workplace injuries. Before a company can pay out on a claim, adjusters must review the details and determine the appropriate level of compensation—a process that typically requires many hours of manual work.
To help insurers compensate claimants faster and more accurately, Acrometis has developed an innovative solution that automates up to 75 percent of the claims-handling process. By ingesting medical and insurance data using optical-character recognition (OCR) and analyzing the information using business logic and machine-learning algorithms, the solution can validate the facts of a case—empowering adjusters to work more efficiently.
After winning clients across the United States, Acrometis is setting its sights on further domestic and international growth. To provide the required levels of performance and availability, Acrometis decided to migrate its applications from a private cloud platform to IBM Cloud bare metal servers.
“Our applications run compute-intensive OCR, data discovery and machine-learning processes, and the performance of our virtual machines is a critical factor in serving our clients effectively,” explains Vinu Varghese, CSO/CTO at Acrometis. “IBM Cloud bare metal servers enable us to quickly spin up new environments with the RAM and CPU resources we need to meet those requirements.”
To comply with its regulators, Acrometis must securely retain its data for several years after it processes a claim. With more than 50 TB of historical data to migrate in addition to its applications, the company wanted a secure and cost-effective approach to long-term storage.
“Because of the sheer volume of data we needed to migrate, it simply wasn’t practical to run the process over our network,” explains Varghese. “To solve the challenge, we looked for a data migration and storage solution that could meet our requirements around speed, security and cost.”