EDMA Insights/DE Support Matrix - Meeting Discussion
- Phuong Nguyen (Unlicensed)
Analytics
Meeting Date | Apr 18, 2023 |
|---|---|
Participants |
|
Purpose | The intention of this meeting is to share the AS-IS status on the current support matrix between Insights and Data Engineering and agree on the direction to move forward. The agreement will come from all Insights TLs, Head of Data Engineering and Data Ops and VP of EDMA. |
Meeting Outcome | This meeting will be successful if
This page will cover
This page won’t cover
|
Key Discussion Points |
|
Problem Statement
We don’t have any documentation to provide all EDMA Insights and Data Engineering team members who manage a particular area to support building Insights dashboards and other Insights initiatives.
It may create a business risk regarding resources, such as resignation or long sick leaves.
Besides resource risk, without this documentation, our team members do not have transparency on who is doing what, which may cause duplication in their work. Transparency is one of the three pillars of empiricism in Scrum. From there, team members all trust each other, and they have the courage to keep each other abreast of good and bad news. Everyone strives and collectively collaborates for the common organizational objective, and no one has any hidden agenda.
The current support matrix can be located at EDMA Support Matrix - version 1
Recommendation
RECOMMENDATION 1
Data Ops will manage all data layers from the Insights team with Primary/Backup Support.
Even though there are some dependencies, this model can help
The query will be written optimally, resulting in saving time and cost in Snowflake.
The insights team will have more time to enhance their current dashboards and implement other initiatives.
We can also overcome the dependency challenges by providing an early head-up to the team for further planning.
RECOMMENDATIOn 2
Data Ops will manage all major data layers/dashboards from the Insights team with Primary/Backup Support. Insights team members can manage the minor ones.
This model can help
The heavy query will be written in an optimised way resulting in saving time and cost in Snowflake
Insights team will have more time to enhance their current dashboards and implement other initiatives
To do this, we need to define the definition of minor and major ones. It may be suitable when our bandwidth in DataOps is high.
Next Steps
Developing TO-BE Support Matrix
Having a plan to put it in practice gradually.