The subject of her research is the history of Soviet prisoners of war during the Second World War. She focuses in particular on the Nazi camp system established in the territory of Ukraine and on the multinational composition of Red Army prisoners. Particularly valuable for this study are the investigative files held by the Central Office of the State Justice Administrations for the Investigation of National Socialist Crimes in Ludwigsburg. These files include materials related to the camps for POWs in Ukraine.

In total, she has obtained copies of more than 800 case files. Processing these archival materials is an extremely laborious task. To facilitate this work, she has begun experimenting with generative artificial intelligence. At present, AI tools can recognize German text in scanned archival documents (PDF and JPEG formats), translate it into Ukrainian, and filter it based on specific keywords or tags. However, the functionality of current GPT-based tools is limited to processing about 20–30 files at a time, while a single archival case may contain at least 200 files.

She is currently developing a scalable automated processing pipeline in Python with the following main features:

• • Bulk retrieval of documents from nested folder structures
• • Keyword/tag search in each file (in German)
• • Metadata extraction (witness name, date, camp location, subject)
• • Document tagging
• • Structured output (CSV or database format)

The integration of generative AI into historical research not only accelerates the processing of large archival collections but also opens new methodological perspectives for analyzing complex historical sources. By combining human expertise with AI-assisted text recognition and data structuring, this project demonstrates how digital tools can enhance the study of difficult historical subjects such as wartime captivity and memory.