Keynote Speaker
Dr. Stefan Tai, Professor

Dr. Stefan Tai, Professor

Technische Universität Berlin, Germany
Speech Title: Trustworthy Data On-chaining in Blockchain-based Decentralized Applications

Abstract: Blockchains guarantee availability, immutability and transparency of transaction records in distrusted, multi-party application contexts. However, data of interest often originates and is held blockchain-externally and, especially, when such data is confidential, the question arises of how to on-chain data from those external sources without revealing any confidential information. We report on our work on ZoKrates und using zero-knowledge proofs (ZKPs) to realize trustworthy data on-chaining with verifiable off-chain computations: All complex, expensive, and data-intensive computations are executed blockchain-externally, and only select results, commitments, and proofs of off-chain computations are stored on-chain, linked with smart contracts. Doing so, a powerful platform is provided to design and operate innovative business applications. We exemplify this approach for designing energy grids for privacy-enhanced energy exchange, for secure federated learning systems in an IoT health context, and for verifiable carbon accounting in supply chains.


Biography: Stefan Tai is Full Professor and Head Chair of Information Systems Engineering at TU Berlin, Germany (2014-present). Prior to that, he was a Full Professor at the Karlsruhe Institute of Technology (2007-2014) and a Research Staff Member at IBM Research in New York, USA (1999-2007). He also held concurrent posts as Director of research labs focusing on technology transfer and is a member of corporate supervisory and advisory boards. He earned his Ph.D. in Computer Science from TU Berlin in 1999.
Stefan specializes on next-generation, platform-based distributed software systems that meet complex system qualities. Platforms of interest include cloud platforms, especially serverless platforms, and blockchain networks, and their meaningful combination and interplay.