Sangam

This is the chatbot interface
This is the backend running

🚀 Inspiration We noticed a gap in how students and professors connect in universities. Cold emails and outdated portals slow down collaboration. We wanted to build an AI chatbot that makes academic networking seamless and personalized.

🤖 What it does Sangam is a smart chatbot that allows students to find like-minded peers and professors to find the best-suited RAs/TAs. It recommends the ideal tech stack for a professor’s project, and once approved, shows the top 5 student matches based on skills and interests.

🛠️ How we built it We used FAST API and PostgresSQL for the backend, streamlit for the frontend.

🧱 Challenges we ran into Designing accurate student-professor matching logic Defining project descriptions in a structured, AI-parsable way Ensuring relevance in tech stack recommendations Handling conversational edge cases in the chatbot flow

🏆 Accomplishments that we're proud of Built a fully functional chatbot with both student and professor modes Integrated AI-based recommendations successfully Created a ranked matching engine for RA/TA recruitment Achieved end-to-end working prototype in limited time

📚 What we learned How to build intelligent chatbots with meaningful context awareness Improving UX in conversation-driven apps Designing matching systems based on skills and interests Balancing technical accuracy with natural user flow

🔮 What's next for Sangam Add a recommendation feedback loop to improve match accuracy Integrate with university authentication systems (SSO) Build a dashboard for professors to manage applicants Extend to industry mentors for internships and capstone projects