About GE Healthcare
GE Healthcare’s Edison AI platform is designed to improve patient outcomes and increase access to care. Edison applications can integrate and assimilate data from disparate sources, and apply analytics or advanced algorithms to generate clinical, operational, and financial insights.
Every patient’s journey involves vast amounts of data. Edison AI converts this ocean of information into actionable insights using analytics, machine learning, deep learning, and Artificial Intelligence (AI). These insights help increase productivity, prioritize workflows, reduce rework, and deliver the most personalized patient care possible.
The Edison AI Workbench helps data scientists to annotate data, train models, and package the data to be deployed into a variety of devices and scanners.
The Edison AI Workbench was originally deployed on AWS cloud, where it makes use of all available AWS resources, to provide a seamless experience to those developing Workbench solutions.
Yet, this cloud architecture ran up against a key challenge in healthcare. Sandeep Lakshmipathy, Director of Engineering for Edison AI sums up the issue: “When we took the Edison Workbench solution to our research customers, they said, ‘This is great. We really like the features and we want these tools, but can we have this Workbench on-premises’?”
The Edison team started evaluating AWS dependencies, along with alternatives capable of being deployed in customer datacenters. During the evaluation, they found DynamoDB was a core component of the solution. The team had two choices: rewrite the Edison Workbench to run against a different datastore, or find a database compatible with DynamoDB.
The team recognized the challenges involved. First, porting a cloud asset to run on-premises is a non-trivial activity, involving specific skill sets and time-to-market considerations. Additionally, the team would no longer be able to perform the continuous delivery practices associated with cloud applications. Instead, they would need to plan for periodic releases as ISO disk images, while keeping code bases synchronized between cloud and on-premises versions. Thus, maintaining a consistent database layer between cloud and on-premises releases was vital to the team’s long-term success.
“The ScyllaDB team helped to integrate with DynamoDB streams, and, overall, partnered effectively.”
– Sandeep Lakshmipathy, Director of Engineering, GE Healthcare
During their search, the Edison team discovered ScyllaDB’s Project Alternator, a DynamoDB-compatible API. Project Alternator enables ScyllaDB to run in hybrid topologies with DynamoDB.
A quick proof-of-concept was all it took to convince the Edison team. “Without changing much, and while keeping the interfaces the same, we migrated the Workbench from AWS cloud to an on-premises solution,” explained Lakshmipathy. “Because of ScyllaDB’s API compatibility, a lot of the code remained the same.”
As an added benefit, ScyllaDB’s Kubernetes support helped with porting microservices and accelerated deployment.
ScyllaDB’s Project Alternator accelerated the timeline in which the Edison team could take the Workbench solution back to their customers to precisely address the requested use case: having the Workbench run in the hospitals’ networks.
ScyllaDB support was always available, providing developer-level support, quick turn-arounds for nightly builds, along with regular check-ins with technical teams. According to Lakshmipathy, “the ScyllaDB team helped to integrate with DynamoDB streams, and, overall, partnered effectively to create the on-premises Edison Workbench solution.”