CampusCare AI

Inspiration

College students deal with many small but important tasks every day—tracking deadlines, understanding college information, reporting campus problems, and deciding what needs attention first. These tasks are often repetitive, scattered across different platforms, and require students to manually search for information or contact the right person.

We wanted to build something that goes beyond a traditional chatbot. Instead of simply answering questions, CampusCare AI acts as an autonomous assistant that can understand a student's request, decide what action is needed, use the appropriate tool, and provide a useful result.

Our goal is simple: let students focus on their studies while an AI agent handles repetitive campus tasks.

What it does

CampusCare AI is an AI-powered campus assistant designed to help students with everyday college activities.

The agent can:

📚 Answer questions using available campus information
📅 Create and manage student tasks and deadlines
⚡ Prioritize tasks based on urgency and importance
🚨 Create and categorize campus complaints
📍 Associate complaints with locations or facilities
🔄 Track the status of submitted complaints
💬 Interact with students through a simple conversational interface

The important difference is that the system is agent-based. The AI doesn't have to rely on a fixed response for every request. It can determine which tool is appropriate and perform the corresponding action.

How We Built It

CampusCare AI uses AWS Strands Agents as the core agent framework.

The application consists of:

React.js — user interface
Python — agent/backend logic
AWS Strands Agents — autonomous AI agent
MongoDB — storing tasks, complaints, and relevant campus information
AWS services — cloud infrastructure and AI capabilities

The basic workflow is:

Student
   ↓
React Web Interface
   ↓
AI Agent
   ↓
Understand the request
   ↓
Select the appropriate tool
   ↓
Execute the action
   ↓
Store / retrieve information
   ↓
Return the result to the student

For example, when a student reports a broken projector, the agent can identify the issue, categorize the complaint, determine its priority, create the complaint, and return the complaint status to the student.

Agent Tools

Instead of putting every operation directly inside the AI prompt, we designed separate tools that the agent can use when necessary.

Examples include:

Task Tool — create and manage tasks
Priority Tool — determine task urgency
Complaint Tool — create and update campus complaints
Information Tool — retrieve campus-related information

This tool-based architecture makes the agent more useful and allows new capabilities to be added without redesigning the entire application.

What We Learned

This project helped us understand the difference between a traditional AI chatbot and an AI agent.

We learned how an agent can:

Understand a natural-language request
Decide what action is required
Select an appropriate tool
Execute that tool
Use the result to generate a response

We also learned how to connect an AI agent with application logic, databases, and web interfaces rather than treating AI as an isolated chatbot.

Most importantly, we learned that good agent design is not only about making the model more capable. It is also about giving the agent the right tools, clear boundaries, reliable data, and meaningful actions.

Challenges

One of the biggest challenges was designing the system so that the AI could perform useful actions instead of simply generating text.

We also had to think about:

Designing reliable tools for the agent
Handling incorrect or incomplete user requests
Making sure database operations are safe
Determining appropriate task priorities
Connecting the agent with the web application
Keeping the interface simple for students
Testing different types of natural-language requests

Another challenge was learning AWS Strands Agents and understanding how agent workflows differ from conventional web application development.

Future Improvements

CampusCare AI can be extended beyond the initial prototype.

Future versions could include:

Integration with college ERP systems
Email and notification integration
Voice-based interaction
Automatic document understanding
Faculty and administrator dashboards
Multi-agent collaboration
Real-time campus announcements
Analytics for frequently reported campus problems

Why It Matters

CampusCare AI is designed around a simple idea:

AI should not only tell people what to do—it should help them get things done.

By handling repetitive tasks while leaving important decisions to people, CampusCare AI aims to make everyday college life simpler, faster, and more organized.