EduAgent AI

Inspiration

As a BTech Data Science student, managing tight assignment schedules and college timetables is tough. We wanted to build a next-generation AI Agent that doesn't just chat, but actively schedules and alerts students about their coursework using Google's latest frameworks.

What it does

EduAgent AI uses Gemini 3.5 to process natural language inputs from students (e.g., "Schedule a Java project reminder for tomorrow at 4 PM"). It automatically parses the date, connects with Google Cloud Firestore database, and structures an autonomous reminder queue.

How we built it

We built the core agent logic using Python and the official Google ADK (Agent Development Kit). The backend is integrated with Gemini 3.5 models via Vertex AI, and deployed using Google Cloud infrastructure to ensure scalability.

Challenges we faced

Since we are learning Python in college, integrating multiple cloud components and managing API authentications within tight deadlines was challenging, but it helped us learn modern backend workflows.

What we learned

We learned how to connect LLMs with custom external databases via functional tools and discovered how Google ADK makes building autonomous systems much more streamlined.

Built With

• gemini-3.5
• google-adk
• google-cloud
• python