Friedrichshafen, Baden-Württemberg, Germany
Computer scientist (PhD), AI scientist and foundation-model architecture researcher with long-standing work in neural computation, machine learning, evolutionary optimisation, industrial AI and complex technology systems. Career spanning early neural-network and evolutionary-computation research, industrial AI applications, Siemens global-headquarters-level complex technology programme leadership, and current architectural research for large language models and multimodal foundation models. Current research focuses on layer-internal associative memory for foundation models: introducing explicit associative memory, competition/inhibition, iterative settling and controllable read/write mechanisms inside Transformer layers, so that memory becomes part of the model’s forward computation rather than an external RAG system, context extension or memory token. Long-term validation targets include multimodal and embodied settings where models must bind perception, language, action history and later queries into reusable internal episodic memory. Early work at Siemens Corporate Research in Germany and the United States covered neural networks, genetic algorithms, expert systems, parallel computing and industrial optimisation. Results led to a German patent, international publications, an AI book, and applications in chip partitioning, surface-acoustic-wave filter optimisation, steel-cooling optimisation and refuelling optimisation for pressurised-water reactors. Later served at Siemens global headquarters as a senior international project and programme lead, managing large and complex IT projects and programmes, cross-border carve-outs, multi-country transformations, major infrastructure migrations and global deployments. Most projects built new IT infrastructures and operating models for global customers in the context of M&A, carve-outs or major organisational transformation. Also entered Siemens’ corporate IT audit function, mandated at CEO level, reviewing global business units and central corporate functions. This creates an unusual senior profile combining frontier AI research, industrial AI validation, enterprise-scale IT and digital-systems governance, corporate technology audit, international programme execution, organisational leadership, risk judgement and strategic implementation.
Continues to advance frontier AI research, technological innovation and complex technology projects. Current focus is the layer-internal associative memory architecture direction for large language models and multimodal foundation models, moving the research path from controlled vision proof-of-concept toward 100M–300M-parameter foundation-model pilot validation planning. Representative work includes: • Originated and advanced the Lateral Associative Transformer architecture route. • Designed and tested DS-LHT / Phase-0 prototypes and the subsequent operator suite. • Established matched controls, ablation modes, run ledgers and causal guards. • Designed the PCam / CIFAR visual-validation track, multimodal retrofit path and foundation-model pilot route. • Planned the research roadmap, team architecture and technical-governance framework required for a future frontier AI lab. Other representative work includes AI-based cell-diagnostics technology research, AI approaches for energy-storage management systems, and complex technology and governance projects such as CHESS / Atos global access, the C&A Bank migration to Bank11, and ABSC project-management-system development.
Teaches artificial intelligence at university level, covering machine learning, deep learning, generative AI, model evaluation and application development; organises courses and practical projects, helping students build systematic AI understanding and experimental capability.
Served for many years in Siemens global-headquarters-level project and operations environments as a senior international project and programme lead, responsible for large cross-border technology projects, programme governance, carve-outs, major infrastructure migrations, server-based-computing architecture cross-project topics and global deployments. Responsibilities covered international project management, senior internal consulting, global-operations project management, carve-out topic leadership, cross-project team leadership and PM@Siemens project-management-system assessments, with a focus on strategic execution in highly complex, multi-country, multi-team and multi-stakeholder environments and full responsibility for schedule, quality, budget, risk and stakeholder satisfaction. Representative projects included: LAN infrastructure migration for approximately 250,000 employees across the EMEA region; Siemens Enterprise Services Europe carve-out; Nokia Siemens Networks global carve-out; and Fujitsu Siemens-related carve-out and service migration. Projects involved phased deployments across more than 30 and up to 80 countries, migration scenarios affecting tens of thousands to hundreds of thousands of employees, and programme budgets ranging from several million euros to more than EUR 10M (German budget only, excluding investments), with delivery quality and stakeholder satisfaction as core success criteria. This experience directly supports the ability to build and run a frontier AI research organisation in high-uncertainty, high-complexity, multi-team environments
Worked in Siemens’ Corporate Planning and Development department as an information technology auditor, reviewing global business units and central corporate functions. Work focused on organisational and information-processing audits, enterprise information-systems audits, assessment of the economic efficiency and future viability of IT investments, evaluation of business-process support through information systems, and development of audit knowledge bases for information and communications infrastructure. Selected for a highly screened corporate-headquarters audit system; globally, only approximately 50 people were admitted, and the function was directly authorised by former Siemens CEO Heinrich von Pierer. This phase developed the ability to evaluate large enterprise technology systems from the perspectives of corporate governance, technology-investment efficiency, business processes, information-systems risk and organisational structure, while strengthening systematic judgement of complex international business structures, information-systems environments and governance issues.
Worked in Siemens’ corporate research system in Germany and the United States on AI and advanced computing R&D, with emphasis on neural networks, genetic algorithms, expert systems, parallel computing and industrial optimisation methods. Representative work included: • Built an expert system for physical mapping / chip partitioning of digital designs. • Served as an important contributor to a European ESPRIT parallel-computing research programme, including application implementation, parallel simulation-program development, finite-element simulation evaluation and work related to the EDS parallel-computing environment. • Researched neural-network learning algorithms based on evolution strategies, combined neural networks with genetic algorithms, developed a new machine-learning training method and obtained a German patent in 1993. • Built mathematical models of genetic operators and applied neural-network and genetic-algorithm methods to image recognition, surface-acoustic-wave filter optimisation, steel-cooling optimisation, refuelling optimisation for pressurised-water reactors, and chip partitioning for semiconductor and digital-circuit design.
In parallel with full-time Siemens work, served as an adjunct lecturer at the University of Hanover, teaching artificial intelligence and computer architecture.