Dataset Lineage & Data Provenance - Interactive Visualization

Dataset lineage answers the critical question: which raw data produced this model? A full lineage DAG tracks data from its raw sources through ingestion jobs, preprocessing pipelines, feature stores, and versioned training datasets into the final model artifact. Tools like DVC, Delta Lake, and Apache Atlas provide lineage tracking. Without lineage, debugging a model degradation is nearly impossible - you cannot tell if the issue is a code change or a data change.

• Click any node to highlight its complete upstream dependency chain
• Each node shows the data system, size, and time since last update
• Trace the path from Raw Events through Spark ingestion, PySpark preprocessing, Feast feature store to the training dataset
• Full lineage enables rollback: if a model degrades, you can re-train on the previous data version

Part of the EngineersOfAI Interactive 3D - free interactive visualizations covering every major concept in machine learning and AI engineering. Hover any element for a plain-English explanation. No code required.