Abstract:
In this project, we propose the development of LLM-based chatbot to enhance hospital ap- pointment booking systems and general query handling processes. Leveraging advanced Large Language Models (LLMs) such as Grok, GPT. The chatbot aims to automate and streamline communication between hospitals and patients. This system connects to the hospital man- agement systems and external APIs which enables them to provide realtime updates on the doctor’s schedule, available appointments, and on specific services offered in each department. The project aims at designing an interface (UI) that is flexible and can be integrated onto to any hospital’s website or system. Human-like conversation is achieved using natural language understanding (NLU) and dialogue management methods which allows for improved context recognition during interaction, far more sophisticated than traditional response generation in chatbots, call centers, and other automated systems. A key highlight of this solution is it’s scalability enabling them to serve several hospitals irrespective of their size or infrastructural complexity. It uses Grok, a powerful proprietary language model, to build a smart chatbot de- signed for hospital communication. While Grok isn’t open-source, it offers strong reasoning skills and performs well in handling tasks like answering medical questions and scheduling ap- pointments. We’ve integrated it with a modern backend using advanced frameworks and secure APIs, ensuring everything aligns with healthcare standards like HIPAA and GDPR. This setup helps maintain data privacy, makes the system flexible, and allows it to work smoothly with existing hospital systems. The final deliverables include a fully functional chatbot, clear tech- nical documentation, and open-access code samples available on platforms like GitHub. By automating routine tasks, responding to patient queries efficiently, and improving the overall experience, this project brings a practical, cost-effective upgrade to how hospitals communicate with their patients.
