Alexander Z.
Principal Data Scientist in Ai Products @Wargaming
Signup · Get unlimited contacts
WORK HISTORY
Principal Data Scientist in Ai Products @Wargaming
Improving bot AI across multiple large-scale session-based vehicular combat games. Moving from \"default bots\" to advanced, adaptive and performant AI agents (opponents and squad mates) that feel human without breaking fairness or server budget.• Technical leadership for training low-level decision making, e.g, steering control, with Reinforcement Learning and Imitation Learning • Overseeing team members training models for high-level decision making, e.g, endpoint position prediction, with advanced supervised learning architectures
EDUCATION
Technical University of Munich
Computer Science
Ludwig-Maximilians-Universität München
Bachelor of Science - BS, Computational Linguistics + Computer Science
Technische Universität Darmstadt
Master of Science - MS, Autonomous Systems
ABOUT ALEXANDER Z.
Currently working on Research & Development for AI in Video Games, mainly in the areas of Reinforcement Learning and Imitation Learning. Worked previously also on Natural Language Processing and classical Deep Learning problems. • Reinforcement Learning and Imitation Learning: 7 years of experience, most recently with video game bot training, previously with participation in the autonomous driving research project \"KI Wissen\" and development of a goal-oriented chatbot• Natural Language Processing: 10 years of experience, most recently retrieval augmented generation with LLMs, entity extraction and alignment for job matching, dialog management and intent recognition for a smart speaker, and sentiment analysis for webpages• Applying state-of-the-art technology to customer projects and supporting the customer from requirement definition, over implementation and testing, up to deployment and maintenance• Technical leadership for all mentioned projects• Organizational leadership for multiple machine learning specialists and management of their competence growth by organizing knowledge transfers, competence coachings and career plans together with themTechnologies worked on:• Programming Languages: Python, Java, Groovy• Machine Learning Libraries: Pettingzoo, Gymnasium, Unity MLAgents, CleanRL, stable-baselines3, imitation, d3rlpy, pytorch, Keras, scikit-learn, TensorFlow, MLFlow, cleanlab, ClearML• NLP Libraries: HuggingFace, gensim, spaCy, NLTK, Word2Vec, FastText, TextAttack, snorkel• Databases: ElasticSearch• Version Control: Git, SVN• Agile Methods: Scrum, Kanban, SAFe
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.