Getting out of the “technical servitude” in the Data Systems
Getting out of the “technical servitude” in the Data Systems For thirty years, data has been built on a succession of transformation tools : ETL, ELT, with increasingly complex jobs, therefore increasingly long pipelines and obviously a multitude of data visualization tools. However, the “technical servitude” does not so much come from the multiplicity of solutions as from their opacity: when no one really knows how a pipeline works from end to end, when the business rules differ from one tool to another in the data visualization layer without anyone clearly identifying it, then the data loses much of its value. A near-unanimous desire is emerging today: to break free from this "servitude" through genuine interoperability. And paradoxically, it is SQL - the language of the past and the horizon of the future - that seems to be becoming the key to building truly interoperable systems. SQL is becoming increasingly important in new architectures....