A series of automated steps that carries data from the systems that create it to the places that use it, cleaning and reshaping it on the way.
Data Pipelines automate the extraction, transformation, and loading (ETL) of raw data into usable formats. In AI systems, they continuously ingest PDFs, remove PII, generate vector embeddings, and update the Vector Database, so the RAG system works from current enterprise knowledge.
See also Vector Database, Data Mobilization and Data Lakehouse.