A few years ago, I worked with an organisation that sells automotive intelligence to streamline the way they got insight from data. I came up with a generic data pipeline to explain to the board how their new data science process could work. It was a hit!
Visuals are a great way to explore a concept and explain a process that could otherwise lose folks along the way.
The key to a good data pipeline is it’s part of an overall process (not shown here) where you know what the problem is, why it’s important to solve it and that data is definitely going to help.
The pipeline focuses on continuous feedback – feedback at every key stage of the process. This could be to the problem owner, other teams, or any other stakeholder to keep them informed and fold their feedback back into the pipeline.
So, here’s my blast from the past – feel free to substitute out Domain Data Science step for other processes that make sense or drop it altogether; whatever works for your situation.