{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/29575ac8814c46d6bb0211ae0214a095\" frameborder=\"0\" width=\"1920\" height=\"1440\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1440,"width":1920,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1440,"thumbnail_width":1920,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/29575ac8814c46d6bb0211ae0214a095-bb56f8aa95af99b5.gif","duration":188.348,"title":"Draft 1 Of AFIT Admission Assistant","description":"This Streamlit application is an AI-powered admission chatbot designed for the Air Force Institute of Technology (AFIT), Kaduna. It helps prospective students get accurate and structured information about AFIT’s admission process, requirements, programmes, fees, and deadlines, using a Retrieval-Augmented Generation (RAG) approach with Groq’s free Llama 3.1 API.\n\nHere’s what it does, in plain terms:\n\nA comprehensive knowledge base is hardcoded into the script — containing official AFIT information such as admission criteria, fees, programme lists, deadlines, contact info, and FAQs. The bot doesn’t “guess” answers; it searches through this structured text for relevant sections using a custom `search_knowledge_base()` function that matches user queries to predefined topic areas.\n\nThe `get_ai_response()` function then passes the user’s question and retrieved context to Groq’s Llama model, generating a clear, friendly, and context-bound response. It’s designed to give accurate, concise, and official** information, refusing to fabricate details when the context doesn’t provide them.\n\nOn the frontend, Streamlit handles all user interaction. The UI features:\n\n1. A sidebar for setup instructions and API key input.\n2. A centered AFIT logo and title using column layout and HTML alignment.\n3. A persistent chat history, displayed in chat-style format with role-based messages.\n4. A chat input box where users type questions.\n5. A footer section crediting the author and linking to GitHub.\n\nIn short, it’s a functional, self-contained admission assistant app that combines structured data retrieval with LLM reasoning — providing a user-friendly interface for AFIT applicants to interact with official admission information in real time."}