Fedor Makovskii

RAG and document assistants, semantic search and embeddings, LLM evaluation (LLM-as-judge, benchmarks), continual learning for small language models, embedding robustness and privacy