Recently I've been delving back into the world of Oracle. Oracle was the first database platform I used. I learned what I know both while at UCA and at Acxiom.

Since then I worked in Oracle for building very large OLAP marketing databases and had the privilege of working with some of the best minds in the industry. Since then I have acquired many more skills and knowledge and experience of other database platforms as well as data management. However, in the last few months I've began working with an org who have a couple of older Oracle deployments. These are traditional "on prem" and they have served parts of their business well for a long time. For reasons having to do with how technology matures over time, this org has a need to identify these processes so they can be migrated to other platforms like GCP or Databricks.

Whatever the end result, the current state is difficult to know and so migrations are difficult because the objective is not to recreate the logic or processes exactly, but to also migrate them to new platforms and reconstitute in those environments in ways that are performant and efficient to their purpose.

Before anything like that can be done one needs to have some kind of strategy. And really, in this case strategy and work will evolve over time. Central to that strategy is to evaluate and unravel what the current processes are doing and to do that I am using SQLMesh, which has proved to be an excellent visualization tool for SQL and Database ETL logic. 

With SQL Mesh I have been able to document to my own understanding how these business processes are consuming then creating new information and data in such a way as to facilitate other members of the business to then offer input into how these ETL processes should change to accommodate the new businesses systems and processes. 

As these ETL processes are more well understood and documented, this becomes the beginning of establishing templates for the new ELT processes which are being built concurrently.