Cloudphysician moved its primary infrastructure from a different cloud provider to Google Cloud at the end of 2023. “Certainly, cost was a factor in our decision to move to Google Cloud, but we also wanted to try Med-PaLM, Vertex AI, and Gemini tools, so it made sense to have our entire infrastructure on Google Cloud,” says Dhruv Sud, Director of Technology at Cloudphysician.

“We’re not looking at diagnosing one disease or figuring out how to read CT scans,” says Dileep. “We’re taking a much wider approach to operational AI, to make doctors and nurses more productive and efficient by enhancing operations. Essentially, we’re removing monotonous tasks by replacing them with AI entirely, so that nurses in hospitals can focus more on patient care.” says Dileep.

With Cloudphysician, the lack of critical care trained nurses in hospitals around the world can also be addressed since tasks like admission and discharge documentation, medication reconciliation skin assessments, MRI pre-checks, and review of care plans can all be done virtually. Cloudphysician already has the largest virtual nursing team in India with more than 100 virtual nurses supporting more than 200+ hospitals.

“We want to partner with hospitals that need a bundled solution that not only has cost-effective tech to connect ICUs but also comes with a layer of virtual nurses that are US-certified to leverage the tech and digitally offload the strained bedside nurses,” says Dileep.

Platform availability is also very important in critical care, which is why the Cloudphysician team uses Kubernetes for horizontal scaling. With Google Kubernetes Engine, the team gets quicker updates with better container orchestration. “Kubernetes updates are faster in Google Kubernetes Engine, and we can deploy nodes in seconds, scaling our infrastructure as our loads spike,” says Dhruv.

The team was given access to Med-PaLM in March 2024, and while some work is still being done to train the model for accuracy, Cloudphysician’s primary use-case for Google Cloud has been through its Google Meet and Google Chat functions.

Separate from the Meet and Chat function, Cloudphysician also uses high-definition cameras with an ongoing video stream of the patient and all of their vitals for monitoring their condition in real time. “We have a separate application called RADAR, which is a mobile and web application that runs concurrently, and in the future, it would be great if Meet could be embedded within RADAR so that we can run everything in a single system,” says Dhruv. Using RADAR alongside Google Cloud, Cloudphysician has supported a 40% reduction in ICU mortality rates within their partner hospital network across India.