[eat_your_potato]

Maybe you can transition to ML-Ops : it's building stuff around Data Science pipeline, and it's better to understand why it needs to be built.