Tassilo J. Klein, Ph.D.

AI research and technology leadership across foundation models, agentic AI, and enterprise systems.

My work connects advanced AI research with real-world systems, from representation learning and language models to agents, simulation, and AI-native products.

---

About

I am an AI research and technology leader with 15+ years of experience spanning foundation models, agentic AI, reinforcement learning and simulation, natural language processing, privacy-preserving learning, continual learning, and multimodal representation learning.

I have built research programs from scientific questions and model design through data, evaluation, and transfer into real-world systems. At SAP, my work included shaping machine learning research for complex, structured environments, building research teams, defining technology agendas, and connecting industrial research with an international academic network.

I have authored 50+ peer-reviewed publications, including five first-author ACL papers, and am a co-inventor on 30+ patents. My research has received more than 5,500 citations.

My academic background includes postdoctoral research at MIT CSAIL and Harvard Medical School, a Ph.D. summa cum laude from the Technical University of Munich, and membership in ELLIS.

LinkedIn · Google Scholar · GitHub · ORCID

---

Current Focus

Area	Focus
Foundation Models	Learning across structured, relational, and multimodal data in complex environments
Agentic AI	Agents, application world models, planning, tool use, and coordinated action
Reinforcement Learning & Simulation	Multi-agent learning, simulation environments, and synthetic experience for decision-making
Learning and Adaptation	Self-supervised learning, contrastive learning, post-training, and continual adaptation
Responsible AI	Privacy, robustness, controllability, safety, and responsible deployment
Representation Learning	Shared representations across language, vision, structured data, and multimodal systems

---

Recent Publications & Updates

[2026.05] - Three papers accepted at the ICML 2026 Workshop on Foundation Models for Structured Data

• A Generative Foundation Model for Heterogeneous Tabular Data
• TableFactory: Generating Semantically Linked Tabular Data via Multi-Agent Behavioral Simulation
• Statistically Indistinguishable, Operationally Distinct: A Formal Barrier for Tabular Foundation Models

[2026.02] - New preprint on how agentic AI reshapes enterprise boundaries

• The Headless Firm: How AI Reshapes Enterprise Boundaries

[2025.05] - Paper accepted at ACL 2025 on controllable language generation

• Contrastive Perplexity for Controlled Generation: An Application in Detoxifying Large Language Models

---

Selected Publications

• Differentially Private Federated Learning: A Client-Level Perspective
  Paper

• Learning to Remember: A Synaptic Plasticity Driven Framework for Continual Learning
  CVPR 2019

• DeepNAT: Deep Convolutional Neural Network for Segmenting Neuroanatomy
  NeuroImage

• Contrastive Self-Supervised Learning for Commonsense Reasoning
  ACL 2020

• Towards Zero-shot Commonsense Reasoning with Self-supervised Refinement of Language Models
  EMNLP 2021

For the complete publication record, see Google Scholar or ORCID.

---

Publication History

[2024.10] - Papers presented at the NeurIPS Workshop on Table Representation Learning

• SALT: Sales Autocompletion Linked Business Tables Dataset
• PORTAL: Scalable Tabular Foundation Models via Content-Specific Tokenization

[2023.07] - Paper published at ACL 2023

• miCSE: Mutual Information Contrastive Learning for Low-shot Sentence Embeddings

[2022.05] - Paper published at ACL 2022

• SCD: Self-Contrastive Decorrelation of Sentence Embeddings

[2021.11] - Paper published at EMNLP 2021

• Towards Zero-shot Commonsense Reasoning with Self-supervised Refinement of Language Models

[2021.11] - Paper published in Findings of EMNLP 2021

• Attention-based Contrastive Learning for Winograd Schemas

[2021.06] - Paper published at IPMI 2021

• Self-supervised Representation Learning for Medical Imaging

[2021.04] - Co-organized the ICML Workshop on Self-Supervised Learning for Reasoning and Perception

[2020.07] - Paper published at ACL 2020

• Contrastive Self-Supervised Learning for Commonsense Reasoning

[2020.02] - Paper published in NeuroImage: Clinical

• Analyzing the Co-localization of Substantia Nigra Hyper-echogenicities and Iron Accumulation in Parkinson’s Disease

[2019.10] - Paper accepted at ICCV 2019

• Budget-Aware Adapters for Multi-Domain Learning

[2019.07] - Paper published at ACL 2019

• Attention Is (not) All You Need for Commonsense Reasoning

[2019.06] - Paper published at CVPR 2019

• Learning to Remember: A Synaptic Plasticity Driven Framework for Continual Learning

[2018.04] - Paper published in NeuroImage

• DeepNAT: Deep Convolutional Neural Network for Segmenting Neuroanatomy

[2017.12] - Preprint on client-level privacy in federated learning

• Differentially Private Federated Learning: A Client-Level Perspective

---

Research Leadership

• Built and led multidisciplinary AI research teams
• Defined research agendas spanning models, data, evaluation, and deployment
• Established international research collaborations connecting industrial challenges with leading universities and AI laboratories
• Translated scientific advances into practical capabilities across research, engineering, and product organizations
• Mentored researchers and Ph.D. students across academia and industry

---

Service & Mentorship

• Reviewer and program committee member for ACL, EMNLP, CVPR, and NeurIPS
• Co-organizer of the ICML Workshop on Self-Supervised Learning for Reasoning and Perception
• Scientific advisor on agentic systems, AI-native products, and applied AI strategy

---

Former Students & Research Mentees

Researchers I have mentored have continued into research, engineering, and faculty-track roles across leading AI laboratories, universities, and technology companies.

• Stefan Lionar, Ph.D. student at the National University of Singapore and Sea AI Lab
• Jan Nikolas Morshuis, doctoral researcher at the University of Tübingen and Tübingen AI Center
• Aiham Taleb, Senior Applied Scientist in Generative AI at Amazon Web Services
• Artur Speiser, AI Research Scientist at Exscientia
• Jannik Wolff, Machine Learning Research Scientist and Research Associate at BIFOLD
• Daniel Dorda, Scientific Software Engineer at uniqFEED and researcher at ETH Zürich
• Max Bain, Research Scientist at Google DeepMind
• Mahdyar Ravanbakhsh, Principal Research Scientist at Zalando
• Oleksiy Ostapenko, Research Scientist at ServiceNow AI Research
• Colin Samplawski, Ph.D. student at UMass Amherst and Advanced Computer Scientist at SRI International
• Mihai M. Puscas, Computer Vision Researcher at SPORTTOTAL.TV
• Sandro Pezzelle, Assistant Professor in Responsible AI at the University of Amsterdam
• Frederik Pahde, Research Associate and Ph.D. candidate at Fraunhofer HHI and TU Berlin
• Shailza Jolly, Research Scientist at Amazon Alexa AI
• Denis Dushi, Machine Learning Engineer at Amazon
• Vadim Tschernezki, Ph.D. student in Computer Vision at the University of Oxford
• Robin C. Geyer, Ph.D. student at the Institute for Machine Learning at ETH Zürich

---

For research discussions, advisory work, and technology leadership opportunities, connect with me on LinkedIn or explore my work on Google Scholar.

_{Last updated August 2026}