Stefan Grafberger

Stefan Grafberger

Ph.D. Student

BIFOLD & TU Berlin

Biography

I am a Ph.D. student at BIFOLD and TU Berlin in the DEEM Lab, conducting research at the intersection of data management and machine learning. I mainly publish at conferences like SIGMOD and VLDB.

My Ph.D. advisors are Sebastian Schelter and Paul Groth. I work on responsible data management (also in collaboration with Julia Stoyanovich). I spent the first three years of my Ph.D. at the University of Amsterdam in the Intelligent Data Engineering Lab, before Sebastian transitioned to TU Berlin. Before my Ph.D., I did my masters at TU Munich with Thomas Neumann and Alfons Kemper and focused on databases.

During my studies, I interned with Microsoft GSL, Amazon Research, Oracle Labs, and worked as a research assistant at TU Munich.

News

Recent Publications

All publications

(2024). Towards Query Optimizer as a Service (QOaaS) in a Unified LakeHouse Platform: Can One QO Rule Them All?. Conference on Innovative Data Systems Research (CIDR).

(2024). Automated Provenance-Based Screening of ML Data Preparation Pipelines. Datenbank-Spektrum.

PDF

(2024). Snapcase - Regain Control over Your Predictions with Low-Latency Machine Unlearning. VLDB (demo).

PDF

(2024). Towards Interactively Improving ML Data Preparation Code via “Shadow Pipelines”. Data Management for End-to-End Machine Learning workshop at ACM SIGMOD.

PDF

CV

I am a Ph.D. student at BIFOLD and TU Berlin in the DEEM Lab, conducting research at the intersection of data management and machine learning. I mainly publish at conferences like SIGMOD and VLDB.

My Ph.D. advisors are Sebastian Schelter and Paul Groth. I work on responsible data management (also in collaboration with Julia Stoyanovich). I spent the first three years of my Ph.D. at the University of Amsterdam in the Intelligent Data Engineering Lab, before Sebastian transitioned to TU Berlin. Before my Ph.D., I did my masters at TU Munich with Thomas Neumann and Alfons Kemper and focused on databases.

During my studies, I interned with Microsoft GSL, Amazon Research, Oracle Labs, and worked as a research assistant at TU Munich. I also interned and worked as a working student at TNG Technology Consulting in Munich and worked as a teaching assistant at University of Augsburg.

In the past, I have been working on deequ, a library for ‘unit-testing’ large datasets with Apache Spark, PGX, an in-memory graph analytics framework, and Umbra, a disk-based database with in-memory performance. Currently, I work on mlinspect and mlwhatif. The goal is to diagnose and mitigate robustness and reliability issues in machine learning pipelines.

Contact

I’m reachable via email at grafberger@tu-berlin.de.