- Application of Large Language Models within Learning Technologies
- Retrieval Augmented Generation strategies for document-based learning applications
- Adaptivity of learning applications with Prompt Engineering strategies
- Fine-tuning of Open Source LLMs for specific learning contexts
- Deployment and Application of Open Source LLMs within local networks
- Application of LLMs within web technologies, like the generation of structured data
- Semantic Similarity Search based on embeddings of textual and/or image data and Knowledge Representation
- Natural Language Processing within Learning Technologies
Max Upravitelev studied Philosophy (BA) and Global History (MA) before switching to Computer Science. His previous positions included civic education projects, art projects and work at the German Federal Cultural Foundation.
While pursuing his second master's degree (MSc) at the University of Hagen he worked as a Research Assistant at FAME within Learning Technologies and Image Processing projects. After he completed his thesis on “Utilizing Knowledge Graphs in Outside Knowledge-based Visual Question Answering” he joined the Fraunhofer FOKUS' business unit Future Applications and Media as a Research Associate in 2023. His main research interests are reasoning with numerical and symbolical representations of data from different modalities and its application within web-based technologies.