Dice - BUstle Whisper

Inspiration

The inspiration behind BUstle Whisper stems from the fragmented and often overwhelming nature of campus information. Currently, students struggle to find study partners ("Da zi") on social media due to information overload, outdated posts, and the genuine risk of encountering scammers. Furthermore, course evaluations are often scattered across various disparate platforms, making it difficult for students to make informed academic choices. We wanted to build a centralized "Whisper" station—a safe harbor where students can find the right courses, connect with reliable friends, and even receive mental health support when the pressure gets too high.

What it does

BUstle Whisper is an all-in-one campus companion platform designed to improve student life through transparency and connection:

Smart Course Hub: Displays detailed course and professor information. It features a dedicated message board where students can discuss and rate their experiences.
AI Sentiment Analysis: Utilizing NLP, the system analyzes student comments on the message boards to display an aggregate "Positive vs. Negative" sentiment score. This gives users an instant, data-driven "vibe check" of the course without reading every single comment.
Mental Health Safety Net: The system is designed with empathy. If the AI detects severe anxiety or distress in a user's interactions or comments, it automatically provides a direct link to psychological counseling resources.
"Da zi" (Partner) Matching: A dedicated section for finding study buddies or activity partners. Users can post specific requests and directly contact matched peers, streamlining the socialization process and reducing the noise found on standard social media.

How we built it

We built BUstle Whisper using a modern full-stack approach centered around AI integration:

Frontend: We used Vue.js to create a responsive, interactive, and user-friendly interface.
Backend: Node.js handles our server-side logic and database interactions.
AI & NLP: We integrated the Qwen API to power our core intelligence features. We utilized an NLP model to process user-generated content, allowing us to perform real-time sentiment analysis on course comments and detect potential mental health red flags in text.

Challenges we ran into

AI Nuance & Accuracy: Fine-tuning the NLP model to accurately distinguish between constructive academic criticism and toxic negativity was a significant challenge. We had to ensure the sentiment scores were fair and representative.
Real-time Integration: Seamlessly connecting the Vue frontend with the Node.js backend while handling API calls to Qwen required careful synchronization. We had to ensure that the sentiment analysis didn't slow down the user experience.
Trust & Safety: Designing the "Da zi" system to be open enough for easy connection, yet secure enough to solve the "scammer" problem that inspired the project in the first place.

Accomplishments that we're proud of

Meaningful AI Application: We are proud of successfully integrating the Qwen API to transform raw text into actionable data (Sentiment Scores) for students.
Social Impact (SDGs): We are proud that our project directly supports UN Sustainable Development Goal 3 (Good Health and Well-being) by integrating a mental health alert system for anxious students, and Goal 4 (Quality Education) by providing transparent course evaluations to help students improve their learning paths.

What we learned

Full-Stack Synergy: We deepened our understanding of the Vue.js and Node.js architecture and how to manage state effectively across the stack.
LLM Integration: We learned how to effectively prompt and integrate Large Language Models (LLMs) like Qwen into a practical web application, moving beyond simple chatbots to data analysis tools.
Tech for Good: We realized the importance of combining technical functionality with social responsibility, specifically regarding student mental health.

What's next for Dice - BUstle Whisper

Student Verification System: We plan to implement a strict student ID verification feature to further eliminate scammers in the "Da zi" section, creating a fully trusted community.
Personalized AI Recommendations: We aim to upgrade the AI to recommend specific courses or study partners based on a user's past behavior and expressed interests.
Mobile Application: Developing a dedicated mobile version (iOS/Android) to provide students with better on-the-go access to the platform.