A new dataset of Russian and Tatar toponyms (place names) has been created, containing over 9,600 entries with linguistic, etymological, and coordinate information. This dataset is used to train a hybrid question-answering system that can accurately answer geospatial questions in both languages. The system combines a retriever and an extractive reader, achieving high recall rates on test queries. The authors release their dataset, QA corpus, models, and web demo openly on Hugging Face. --- Why it matters: This research matters to engineers working on geospatial question-answering systems because it provides a new bilingual dataset and hybrid architecture that can be applied to real-world services, such as geocoding and digital humanities projects. The high recall rates achieved by the system demonstrate its potential for improving the accuracy of geospatial QA. Source: https://arxiv.org/abs/2605.05962