Greater Munich Metropolitan Area
Seasoned multi-source/multi-sensor data fusion and "big-data" analysis specialist with 20 years working experience in machine learning, high volume data management and global optimization; as a "tech" as well as on management level with international teams. Specialties: R&D management, project valuation and risk assessment for technology-intensive projects (e.g. with Real Option valuation) or projects with high risk profile; Project management methods (V-Model, SCRUM). Operations research; Optimization Metods (a.a. Particle Swarm and Derandomized Optimization). Machine Learning; Artificial Intelligence - AI & DAI, specifically Support Vector Machines (SVM), Deep Belief Nets (parallelized out to CUDA 4/5/6/7 architecture), Recurrent Networks (G-LSTM); Reinforcement learning for Semi- and Partially observable Markov Settings; Bayesian Modelling and Markov Chain Monte Carlo estimation (MCMC); Sequential Markov Models for Tracking (aka "Particle Filters"). High-Performance and Cluster Computing. Advanced C++ Programming for high-performance/big data applications; GPU Computing.
HEDGE21 combines big data analytics for capital markets with geostrategic assessments and company-specific financial data to provide clients with AI-based digital assistants to make optimal, timely, and cost-effective hedge decisions. HEDGE21 is a financial advisory service provided by ALCAZAR Capital Limited, an asset management company regulated by the Dubai Financial Services Authority.
We are active in 3 application fields, where we try to make sense out of large data volumes and compute decision support, which is somehow "optimal under uncertainty": Financial/Industrial Sector: Algorithm and System Development for "Integrated Currency Risk Management", supporting clients who either invest in instruments, which are denoted in foreign currencies (overlay), or, clients trading goods/commodities in foreign currencies, where the job of the system is to compute optimized leading/lagging decisions and optimizing decisions concerning transactional-, hedge-, cash-flow exposure. Predictive analytics models for cash flow and collections forecasting. Optimization under uncertainty for cash-pool operations; ie stochastic optimal control policy to trim settlements between cash-pools in multiple currencies. Logistics Real-time multi-agent logistics optimization system, which dynamically coordinates a nation wide mutualized logistics network in France, including long haul distribution from producers to peri-urban distribution centers to last mile delivery with merge-in transit. Military Sector: - Signal analysis ELINT software system for a large-scale UAS project. - Algorithmic concept and SW for detection, classification and identification of low-probability of intercept (LPI) radars in Cognitive Electro-Magnetic Warfare (CEW) based on machine learning (recurrent attention networks); - Stochastic optimal control based sensor-cueing method on board an autonomous reconnaissance platform; - Sequential Monte Carlo based method for practical ATR in a UAV/UCAV demonstrator. - Adaptive optimal control algorithm coordinating tasks in a group of UAVs (specifically task allocation, sensor cueing, handover and divert). - Advanced CFAR (constant false alarm rate) component based on factored probabilistic model for passive coherent location sensor system. - Simultaneous Localization and Mapping (SLAM) model with Bayesian model and sequential monte-carlo.
- Advanced Bayesian predictive maintenance model for rail-infrastructure (e.g. switch health status) - Intelligent sensor and sensor data analytics for rail-infrastructure monitoring.
Development, maintenance and operation of a multi-agent/swarm intelligence based, algorithmic trading system for a multi-Billion, US-based Hedge Fund using high-performance computing clusters, including a real-time capable, multi-liquidity provider best-execution subsystem, featuring direct integrations into the major FX dealing banks, liquidity pools (Currenex and Integral) and the interbank reference system EBS.