Anika Schumann

I am a Senior Research Scientist at IBM Research – Zurich with expertise in Explainable AI and AI fault diagnosis. My work addresses two distinct types of faults:

1. Faults in complex systems: Here the task is to advance AI methods for fault detection in large-scale systems, such as power supply networks, smart buildings, wind farms, and IoT systems. The goal is to identify issues in a timely manner, improving both system resilience and operational efficiency.
2. Faults produced by AI models: Here the task is to diagnose and correct errors that arise from AI models themselves, particularly large language models.

An example of this second area is my current work on the Text-to-SQL task, which allows users to interact with complex database systems using natural language queries. In this project, I drive our effort on automatically detecting and correcting errors generated by the AI model, enhancing both system accuracy and user experience.

In addition to this technical role, I have also held several other roles at IBM Research, including Leader of our AI Research Partnerships in Europe, Research Manager for AI and Industries, and Research Strategist. The latter position I have also held at IBM’s TJ Watson Research Center, New York, USA where I evaluated, recommended, and shaped topics and recommendations that define IBM’s strategic directions.

My PhD I received from the Australian National University where I worked on the automated fault diagnosis of large-scale discrete-event systems like telecommunication or power supply networks.

Selected External Activities:

• Expert committee member of the Helmholtz-Incubator “Information & Data Science”

• AI expert round table participant at the Federal Chancellery in Berlin, Germany

• Member of the OECD expert group on task automatability

• Panelist on different topics, e.g.:

  • "Artificial Intelligence - Blessing or curse for humanity?" organized by the Swiss Institute for Global Affairs in Bern
  • "Can AI help address skill shortages?" organized by the OECD as part of the International Conference on AI in Work, Innovation, Productivity and Skills
  • "Responsible AI: Is There Global Convergence?" at the CPDP conference in Brussels
  • "Let's talk about fairness and red line for predictions?" at the "AI & Humanitarian Action" Track, organized by the International Committee of the Red Cross, of the AMLD conference in Lausanne

• Senior Program Committee (SPC) member / PC / Reviewer for: International Joint Conference on Artificial Intelligence (IJCAI), American National Conference on Artificial Intelligence (AAAI), Conference on Neural Information Processing (NeurIPS), European Conference on Artificial Intelligence (ECAI), International Workshop on Principles of Diagnosis (DX), IEEE International Conference on Consumer Communications and Networking (CCNC), International Conference on Connected Vehicles (ICCVE), IEEE Transactions on Computers journal, Pervasive and Mobile Computing journal, IEEE Internet Computing magazine