Cancer remains one of the leading causes of mortality worldwide, and timely access to reliable information is crucial for patients and caregivers. This paper presents CancerCare Bot, an AI-driven chatbot designed to provide instant, accurate, and accessible cancer-related information. The bot is developed using Intel’s Neural Chat LLM v3, fine-tuned with medical resources and trained on multiple cancer-related datasets. The system integrates Natural Language Processing (NLP) to understand user queries and deliver contextual responses. A Flask-based backend enables seamless communication between the AI model and a user-friendly web interface. CancerCare Bot aims to bridge the gap between healthcare professionals and the general public by offering 24/7 assistance, thus improving awareness and early diagnosis efforts.

The development of CancerCare Bot follows a structured pipeline involving data preprocessing, model selection, training, and deployment:

1. Data Collection & Preprocessing
• Gathered cancer-related medical literature, research papers, and FAQs.
• Converted documents into structured embeddings using BGE (Bi-Encoder) embeddings to improve response quality.

2. Model Development
• Used Intel’s Neural Chat LLM v3, a pre-trained large language model fine-tuned for healthcare-related queries.
• Integrated Flask-based backend to handle user requests and AI responses.

3. Web & API Integration
• Developed a web-based chatbot UI using HTML, CSS, and JavaScript.
• Connected the UI with Flask API endpoints to send user queries and receive AI-generated responses.

4. Evaluation & Testing
• Conducted accuracy testing by comparing AI-generated answers with expert-reviewed responses.
• Performed user testing to ensure ease of interaction and response reliability.

The CancerCare Bot was tested across multiple parameters, yielding the following insights:

Response Accuracy: **Achieved 85-90%**accuracy in retrieving relevant cancer-related answers.
User Experience: 92% of test users found the bot helpful and easy to use.
Latency & Speed: The chatbot delivers responses in under 2 seconds, ensuring a seamless experience.
Medical Reliability: The bot provides responses backed by validated cancer research sources, improving user trust.