Love this breakdown! The shift left mindset for reusability is 🔥
I’ve seen teams over optimize it and end up with brittle upstream models that everyone depends on (and nobody dares change). How do you balance reuse with flexibility for edge cases?
Thanks for this comment! I definitely had cases when following rules strictly caused more problems than solved.
I’d say if you have a strong argument why you want to break the rule, you should do it. For example, go ahead and branch the existing model if maintaining it for all edge cases is too hard.
Thanks Oleg for the awesome breakdown, and looking forward to reading the upcoming parts.
From a CDO perspective, I have noticed too many teams get sucked into lengthy 'holy wars' on which is the best modeling methodology.
The philosophy I advocate is 'Pace over Perfection' - pick a model as a starting point knowing you will end up with a hybrid anyways, start developing in collaboration with business stakeholders, and iterate.
Love this breakdown! The shift left mindset for reusability is 🔥
I’ve seen teams over optimize it and end up with brittle upstream models that everyone depends on (and nobody dares change). How do you balance reuse with flexibility for edge cases?
Thanks for this comment! I definitely had cases when following rules strictly caused more problems than solved.
I’d say if you have a strong argument why you want to break the rule, you should do it. For example, go ahead and branch the existing model if maintaining it for all edge cases is too hard.
Respect!
Thanks Oleg for the awesome breakdown, and looking forward to reading the upcoming parts.
From a CDO perspective, I have noticed too many teams get sucked into lengthy 'holy wars' on which is the best modeling methodology.
The philosophy I advocate is 'Pace over Perfection' - pick a model as a starting point knowing you will end up with a hybrid anyways, start developing in collaboration with business stakeholders, and iterate.
Great approach, love it!