gensai.aiChunking means splitting documents into smaller passages, called chunks, before turning them into embeddings and storing them in a vector database. Chunk size and overlap strongly affect how relevant a RAG system is: too short and they lose context, too long and they dilute the information.
In practice at Gensai
At Gensai, chunking is tuned to each type of document, legal texts, technical sheets or transcripts, so the chatbot finds the right passage.