An ML startup developing an autonomous wind turbine inspection platform needed to manage training data, package versioned models for production, and move containerized packages to workhorse machines. We provided end to end expertise in automating and managing all aspects of their ML pipeline in the Azure ecosystem.
Learn More
An ML inspection workflow that we built for a client has unpredictable resource consumption needs - there are frequent "bursts" where we need a multitude of GPUs to process the jobs in a timely manner, but it would be a steep and unnecessary expense to keep these machines running when not in use.
We automated their infrastructure using Azure so that the system will automatically spin up new GPUs to quickly fulfill jobs, and decommission the machines when they are no longer needed.
Learn More