We are building SystemDS, an open source ML system for the end-to-end Data Science lifecycle (integration, cleaning, preparation, training, debugging, serving).

We intend to complement our systems-oriented research by applied research on data management and advanced analytics for specific domains and non-relational data models like time series, graphs, and semi-structured documents.

We aim to investigate language abstractions for data integration, data cleaning, anomaly detection, as well as data validation and debugging for ML workloads.