Intellichat: A Web-Based Domain-Specific Chatbot for Syllabus-Aligned Technical Learning

Abstract Domain-specific academic support is essential for Computer Science and Engineering students, as technical subjects require accurate and syllabus-aligned explanations. Conventional learning resources such as textbooks and faculty consultations can be time-consuming and may not provide instant clarification. To overcome these limitations, this study presents Intellichat, a domain-specific chatbot designed to assist CSE students through real-time conversational learning. The system integrates a web-based interface with a Flask backend and utilizes a Large Language Model through Cohere API for generating intelligent responses. User queries are first processed using a keyword-based domain filtering mechanism to ensure that only Computer Science-related questions are accepted. Relevant prompts are then structured and forwarded to the language model, which produces context-aware academic explanations for subjects such as Data Structures, Operating Systems, DBMS, and Computer Networks. Experimental evaluation using response relevance, accuracy, and latency metrics demonstrates that Intellichat provides fast and syllabus-focused assistance compared to general-purpose chatbots. Keywords: CN Domain-Specific Chatbot, Large Language Models, Computer Science Education, Cohere API, Prompt Engineering, Academic Assistant