• Enterprise Analytics Community

  Unanswered Questions from September 28 Session – Data Governance

  Posted by Leslie Cook (Adm) on September 29, 2023 at 5:03 am

  Hello UAI members, 

  There were a couple of unanswered questions from the September 28 Enterprise Analytics Community Conversation. Please find those questions below with answers from our panelists. Please continue the conversation on data governance! You can use this thread to ask questions to your fellow members. 

  1. How does your analytics data governance interact with data governance around your operational applications like SAP? 
     • Thoughts from @Joseph: If I’m understanding the question correctly, SAP is our billing system. The security is tight as you can imagine. Not only are there rules around access to SAP Hana, but also around the what specifically. My team has generally significant amounts of access simply due to the official numbers, customer info, account assignment (c4c), and billing information. We talk often to them to provide use cases and enabling the architects/engineers/etc. ensure there is correct methodology in place before moving the data to a production state even in the specific application(s) downstream. In short, we are constantly talking to IT, SAP, etc. to ensure the data is used as intended based on the problem we’re solving. Taking a step back, looking at the big picture, understanding the other analytics community groups similar uses helps quite a bit too. In SAP Hana Production language, having too much load of data pulls can literally take down the whole billing system. As mentioned, SAP requires a lot of patience. The best bet is to work to transport that data somewhere else as a mirror copy putting some off hour schedule in place to put the data in a safe space (no harm to their prod). The same logic above applies to ensure we are using the data appropriately working in collaboration to curate the code + sources.
     • Thoughts from @Hannah: We are also an SAP shop. Our data governance appreciates SAP’s data governance! But we’re separate from SAP data governance. We generally try to reveal and support enforcement for SAP data governance needs (where we have SAP data enabled in our production data lake this is possible), and obviously work to support SAP governance. Much of our data quality improvements are resulting in us having to go back into SAP to fix at the source, so we are working together with our SAP teams to ensure that we have good governance end to end. Our general philosophy around working with operational application governance is that we understand we are stuck with whatever governance or capabilities exist in our source systems, BUT, we try to get to a unified feeling in our data lake and analytics governance. This is sometimes easier said than done, so every time we integrate a new system we look at the impacts on existing guidelines and make changes to those guidelines, if necessary. 
     • Thoughts from @Kelly: Our process for data governance between data analytics and enterprise systems are very similar.  We identify system/data owners and have perform approval for access control, user access, data access, and data use case approvals.
  2. You’ve mentioned the Data Lake specifically a few times. Is that where you’re focusing most of your efforts for Data Governance/Cataloging, or are you looking at ALL data in the company? 
     • Thoughts from Joseph Katon:  We’re focused on all data and building the catalog out to go end-to-end. It would be nice to know the sources and who’s application(s) they’re going to. This could include a Power BI report, too as a side note.
     • Thoughts from Hannah Ball:  Yes! One of our taglines in data governance is that if it isn’t in the datalake, it didn’t happen. 😝 In all seriousness though, we are encouraging teams to bring their data into the datalake so it is accessible with enterprise tools and processes (ETL, reporting, data catalog, etc) and makes managing access, data classification, public data requests, etc SO much easier. This also means finding and using a data catalog solution is simpler because we aren’t having to add a bunch of different integrations. That being said, there may be some applications where it doesn’t make sense to bring data into the datalake (ArcGIS is one that comes to mind), so we DO have integrations to those environments.
     • Thoughts from Kelly Laurie: You just need to start on both data lake and data cataloging.  It will be difficult to identify all the data in both scenarios. We started with AMI in lake and are adding additional data sets as use cases present themselves.  On the catalog side we started with existing analytic solutions back tracked, validated, and moved forward.  I see both efforts as related and need to be coordinated but not a single united project.

  I hope everyone has a wonderful weekend! 

  Cheers, 

       Leslie 

  ——————————
  Leslie Cook
  Membership, Engagement, & Training Manager
  Utility Analytics Institute (UAI)
  719-203-8650, lcook@utilityanalytics.com
  ——————————

  Leslie Cook (Adm) replied 1 year, 1 month ago 1 Member · 0 Replies
• 0 Replies

Log In to Reply