Live Resource Monitoring and Management
Deploying AI applications and experiences is a complex endeavour and requires many moving parts to work together in harmony. Between the AI apps that are deployed and the ongoing data science research it takes to support and continuously improve those deployments there are a great deal of resources to manage and monitor
With cnvrg you are empowered to leverage ALL of your resources, whether they are on AWS, Azure, GCP or on premise. This allows organizations to choose the best resources for development, testing and deployment across any environment from a single platform. This would traditionally place a huge burden on IT and create a friction point between IT and the organization trying to push the AI initiatives forward. By using cnvrg Engineers and Scientists have full access to all resources while IT can monitor and manage from a single pane of glass, that is the cnvrg MLOps Dashboard.
The dashboard is updated in real time to show the most current view of the entire infrastructure and utilization/allocation details for all resources. Users are able to drill down into the resource allocations AND utilization across all clouds and on premise resources. Examine CPU, GPU and Memory for every running Notebook, Deployed Application, Experiment and process in a centralized location with the capability to inspect those jobs and stop them if needed to free up unused resources for others to leverage.
The chart shows a high level aggregation of all resources at each time interval while also providing a hover tooltip that displays the individual contributions to the overall utilization. The user can drill down into the app from here to investigate further.
Selecting a deployed model gives you detailed insights around the endpoint metrics like the number of requests coming in as well as the response latency of your model.
All data can be exported and used in your prefered BI tool to examine the data in more detail and produce reports. This is particularly useful when forecasting future growth and usage planning as well as in chargeback models for easy allocation of spend.
This is just one of the many features that makes cnvrg the #1 AI Platform for enterprise users engaged in research and development of the next generation of AI applications.
