I have a customer scenario where I need to use Operational data for Data Science & Machine Learning purposes.

As a customer that has assets in locations that have few personnel that are capable of maintaining software on site, but need to create displays, reports, and troubleshoot operational issues using data from several different small automation systems. I need a low to zero footprint offering that is always up-to-date offered on a subscription basis so that I can enable visualization and analytical capabilities on my remote sites.

As an company that uses PI Systems across our enterprise, we need the ability to aggregate data across on premises PI Systems located at our sites.

Today, we often use another "Enterprise" level PI Server to perform this aggregation and 'corporate view'. but in the future, we would like to be able to aggregate this data in a cloud based Software as a Service (SaaS) offering hosted by OSIsoft.

This would enable customers to largely replace their central/corporate PI Servers with OSIsoft Cloud Services (OCS).

We don't have the resources or skills to setup, administer, and manage our own PI System. We would just like to be able to use a hosted system delivered as Software as a Service (SaaS).

Leverage the OSIsoft Cloud Services as a backup and recovery mechanism for site PI Systems to help with disaster recovery (ex: storm response) and/or business continuity.

Whilst OCS' SDS supports storing structures that include nested attributes, OCS Trending currently doesn't support that functionality.

The workaround is to flatten the nested structure to a simple type, and get the OCS Trending to work.