{"id":8325,"date":"2025-11-10T21:00:08","date_gmt":"2025-11-10T13:00:08","guid":{"rendered":"https:\/\/people.utm.my\/shahabuddin\/?p=8325"},"modified":"2025-11-11T09:56:38","modified_gmt":"2025-11-11T01:56:38","slug":"how-i-built-a-custom-ai-chatbot-for-my-university-courses","status":"publish","type":"post","link":"https:\/\/people.utm.my\/shahabuddin\/?p=8325","title":{"rendered":"How I Built a Custom AI Chatbot for My University Courses"},"content":{"rendered":"\n

It\u2019s 2 AM. A student is staring at a folder full of lecture slides, project guidelines, and 80-page PDFs. They have a specific question: “What are the submission requirements for Project Phase 1?” They know the answer is somewhere<\/em> in those files, but where?<\/p>\n\n\n\n

We\u2019ve all been there. As an educator, I see this problem all the time. Students have the materials, but information retrieval<\/strong> is a huge, unaddressed challenge.<\/p>\n\n\n\n

So, I asked myself: “What if I could build a 24\/7 AI teaching assistant for each of my courses? One that only<\/em> knows about that specific course’s materials?”<\/p>\n\n\n\n

That’s exactly what I did. The result is a lightweight, custom-built, multi-course AI chatbot. Here’s how I built it.<\/p>\n\n\n\n

The “Aha!” Moment: Why Not Just Use ChatGPT?<\/h3>\n\n\n\n

The first question is obvious: “Why not just use an off-the-shelf AI?”<\/p>\n\n\n\n

The simple answer: No public AI knows about my private course materials.<\/strong> I can’t ask it about the “SBEG3163 System Analysis” syllabus or the “MBEX1013 Research Methodology” project guide.<\/p>\n\n\n\n

The solution is a framework called Retrieval-Augmented Generation (RAG)<\/strong>.<\/p>\n\n\n\n

In plain English, RAG means we “teach” the AI by giving it a cheat sheet before<\/em> it answers. The process is simple:<\/p>\n\n\n\n

    \n
  1. Retrieve:<\/strong> When a student asks a question, we first search our own<\/em> documents (the cheat sheet) for the most relevant text.<\/li>\n\n\n\n
  2. Augment:<\/strong> We take that relevant text and paste it into a prompt for the AI.<\/li>\n\n\n\n
  3. Generate:<\/strong> We tell the AI, “Using only<\/em> this text I just gave you, answer the student’s question.”<\/li>\n<\/ol>\n\n\n\n

    This gives us the power of a giant AI like Google’s Gemini, but grounded in the specific, factual knowledge of our own private materials.<\/p>\n\n\n\n

    The “Franken-Stack”: Simple, Powerful Tech<\/h3>\n\n\n\n

    I didn’t want a heavy, expensive, complicated system. I wanted something that could run on any basic web host. So, I built a “vanilla” stack that is surprisingly powerful:<\/p>\n\n\n\n