BruinBoard

Overview

BruinBoard is a SvelteKit 2.0 web application that aggregates UCLA Instagram content into an organized platform with calendar events, posts, and AI-generated summaries. It serves as a centralized hub for UCLA students to discover campus events and activities from various Instagram accounts.

Key Features

• Calendar view with event extraction from Instagram posts
• Instagram feed aggregation from multiple UCLA accounts
• AI-generated summaries of events and content
• Category-based filtering system
• Responsive design optimized for mobile devices

Design & Structure

BruinBoard uses SvelteKit with a component-based architecture. The application leverages AI to extract event information from Instagram posts and generate summaries. Data is stored in JSON files for simplicity, with a future plan to migrate to a database.

Key Components

• Calendar Component: Month/Week/Day views for events
• Posts Component: Instagram feed display
• Summaries Component: AI-generated content display
• CategoryFilter Component: Filtering mechanism for content
• Navigation Component: Fixed bottom navigation for mobile

Project Structure

The application follows a clean structure with routes for different views (calendar, posts, summaries), components for UI elements, and AI services for data processing. The design emphasizes UCLA branding with appropriate color schemes and responsive layouts.

// Example AI event extraction from Instagram post
import { openai } from './config';

export async function extractEvents(post) {
  const prompt = `
Extract event details from this Instagram post caption:
---
${post.caption}
---

If there is an event mentioned, provide the following details in JSON format:
{
  "title": "Event title",
  "start_time": "YYYY-MM-DD HH:MM:SS",
  "end_time": "YYYY-MM-DD HH:MM:SS",
  "location": "Location name",
  "description": "Brief description"
}

If no event is mentioned, return { "is_event": false }
`;

  const response = await openai.createChatCompletion({
    model: "gpt-4o",
    messages: [{ role: "user", content: prompt }],
    temperature: 0.2,
    max_tokens: 500
  });

  try {
    const content = response.choices[0].message.content;
    return JSON.parse(content);
  } catch (error) {
    console.error("Failed to parse event data:", error);
    return { is_event: false, error: "Failed to parse response" };
  }
}

Database Schema

BruinBoard currently uses JSON files for data storage, with a planned migration to a database. The schema below represents the planned database structure for storing Instagram accounts, posts, events, and AI-generated summaries.

Technologies

AI Processing Pipeline

BruinBoard uses a sophisticated AI pipeline to extract, process, and consolidate event data from Instagram posts:

API Endpoints

BruinBoard provides API endpoints for accessing and managing data:

Future Directions

Roadmap

• Event date verification system
• Multi-day event support
• Fine-tuning GPT to process events better
• Search functionality across all content
• User accounts with personalized feeds
• Native Mobile Application via Capacitor