Efficient document management is important to optimize operational processes and ensure fast access to the information contained therein. Chatbots are increasingly being used in different sectors, as their implementation opens up promising opportunities to improve access to information. This research presents ChatBanker, a chatbot developed to efficiently manage documents in the banking sector. The main objective is to improve the accessibility and retrieval of document information by using natural language processing (NLP) technologies, the OpenAI API and optical character recognition. The methodology includes segmentation and embedding of PDF content using embeddings, enabling accurate responses to user queries. The results demonstrate ChatBanker’s ability to process multiple documents simultaneously, providing accurate responses based on semantic similarity. ChatBanker achieves a 92% response accuracy and reduces information search time by 70%, demonstrating its effectiveness. The chatbot highlights its potential to significantly improve operational efficiency and user satisfaction in document management tasks. These results underscore ChatBanker’s practical value as an artificial intelligence-driven solution for document-intensive environments.