Inspiration: As students, we experienced firsthand the confusion and challenges of selecting the right courses during our first and second semesters. This inspired us to build something that could make this process easier and more personalized—not just for us, but for all students facing similar decisions.
What It Does: EduPilot is a smart course recommendation tool designed to help students make informed decisions based on their career goals, desired skills, and interests. It provides: 1) Recommended courses tailored to individual goals 2) Key course details including availability, prerequisites, and credit hours 3) Clear explanations on why a course is recommended 4) Additional tips and short notes to guide students on their academic path
How We Built It: We developed a user-friendly interface using Streamlit. The course dataset was curated by parsing IU bulletins and handbooks, extracting details like course descriptions, prerequisites, availability, and credits. For generating intelligent, context-aware recommendations, we integrated a Hugging Face API powered by the DeepSeek-V3 language model.
Challenges we ran into: We faced challenges with to make a the UI dynamic and overcame the challenge by applying customized CSS
What We Learned from EduPilot: 🔧 End-to-End App Development Learned how to design, structure, and deploy a full-stack AI-powered app using Streamlit, combining frontend UI with backend logic.
🧠 Working with LLMs Understood how to integrate a language model API (like Hugging Face) and handle challenges like prompt handling, response timing, and context retention with st.session_state.
🎨 Customizing Streamlit UI Gained hands-on experience in styling and layout using HTML & CSS inside Streamlit, including fixed chat input, sidebar dropdowns, embedded logos, and chat bubble aesthetics.
📊 Embedding External Tools Integrated Power BI dashboards and document links seamlessly within the sidebar, enabling a multi-modal experience for students.
🐞 Debugging Real-world Issues Faced and solved bugs like broken image paths, duplicate prompts, and raw code rendering — improving your debugging and Streamlit workflow skills.
🙋🏻♀️ User-Centric Thinking Designed the chatbot and sidebar features based on student needs, focusing on simplicity, speed, and clarity — a great exercise in UX thinking
Log in or sign up for Devpost to join the conversation.