An AI-powered feed ranking system that aggregates and ranks articles from multiple sources using semantic relevance, recency, and popularity scores.

🌐 Live Demo: https://app-production-075c.up.railway.app/

• Multi-Source Aggregation: Fetches articles from Hacker News, Dev.to, and Reddit (with authentication)
• AI-Powered Summaries: Generates concise, contextual summaries using OpenAI's GPT-4o-mini
• Semantic Ranking: Uses OpenAI embeddings to compute relevance scores based on semantic similarity
• Score Breakdown Visualization: Visual progress bars showing relevance, recency, and popularity scores
• Modern UI: Beautiful, responsive interface with smooth animations and transitions
• Real-time Search: Fast, parallel fetching from multiple sources
• Source Badges: Color-coded badges indicating the source of each article

SignalFeed is a monorepo built with:

• Frontend: Next.js 16 with React 19, Tailwind CSS, and GraphQL
• Backend: Apollo Server (GraphQL) with TypeScript
• AI: OpenAI API for embeddings and summaries
• Package Manager: pnpm workspaces

signalfeed/
├── app/                 # Next.js frontend application
│   ├── app/
│   │   ├── graphql/    # GraphQL queries
│   │   ├── lib/        # GraphQL client
│   │   └── page.tsx    # Main UI component
│   └── package.json
├── server/              # GraphQL backend server
│   └── src/
│       ├── schema/     # GraphQL schema and resolvers
│       └── services/
│           ├── dataSources/  # Data source integrations
│           ├── embedder.ts   # OpenAI embedding service
│           └── feedService.ts # Main ranking logic
└── shared/             # Shared TypeScript types
    └── src/types/
        └── feed.ts

• Node.js 18+
• pnpm 10.15.1+
• OpenAI API key

1. Clone the repository

   git clone <repository-url>
   cd signalfeed

2. Install dependencies

   pnpm install

3. Build shared package

   cd shared
   pnpm build
   cd ..

4. Set up environment variables

   Create a .env file in the server/ directory:

   OPENAI_API_KEY=your_openai_api_key_here

5. Start the development servers

   In separate terminals:

   # Terminal 1: Start GraphQL server
   cd server
   pnpm dev

   # Terminal 2: Start Next.js app
   cd app
   pnpm dev

6. Open your browser

   • Frontend: http://localhost:3000
   • GraphQL Playground: http://localhost:4000

Each article is scored using a weighted combination of three factors:

1. Relevance Score (60%): Semantic similarity between the search topic and article title using OpenAI embeddings
2. Recency Score (25%): Based on publication date
   • Today: 100%
   • Last 7 days: 80%
   • Last 30 days: 50%
   • Last 6 months: 30%
   • Older: 10%
3. Popularity Score (15%): Based on upvotes/points from the source platform

• Hacker News: Uses Algolia search API
• Dev.to: Uses Dev.to API with tag-based search
• Reddit: Requires authentication (currently may be blocked without OAuth)

• Next.js 16: React framework
• React 19: UI library
• Tailwind CSS: Styling
• GraphQL Request: GraphQL client
• TypeScript: Type safety

• Apollo Server: GraphQL server
• OpenAI API: Embeddings and summaries
• TypeScript: Type safety
• Node.js: Runtime

query GetRankedFeed($topic: String!) {
  getRankedFeed(topic: $topic) {
    id
    title
    url
    summary
    score
    relevanceScore
    recencyScore
    popularityScore
    explanation
    source
  }
}

{
  "data": {
    "getRankedFeed": [
      {
        "id": "hn-12345678",
        "title": "Example Article Title",
        "url": "https://example.com/article",
        "summary": "AI-generated summary of the article...",
        "score": 0.85,
        "relevanceScore": 0.92,
        "recencyScore": 0.8,
        "popularityScore": 0.7,
        "explanation": "Relevance: 92% | Recency: 80% | Popularity: 70%",
        "source": "hackernews"
      }
    ]
  }
}

• Skeleton Loaders: Smooth loading states while fetching
• Score Visualization: Progress bars for each scoring factor
• Source Badges: Color-coded badges (HN: orange, Reddit: orange-red, Dev.to: green)
• Copy Link: One-click link copying
• Responsive Design: Works on desktop and mobile
• Smooth Animations: Fade-in animations for feed items

# Build shared package
cd shared && pnpm build && cd ..

# Build server
cd server && pnpm build && cd ..

# Build app
cd app && pnpm build && cd ..

# Server
cd server && pnpm start

# App
cd app && pnpm start

• Reddit API: Reddit may block unauthenticated requests. To use Reddit, you'll need to:
  • Register a Reddit app at https://www.reddit.com/prefs/apps
  • Implement OAuth authentication
  • Or use a proxy service

• Reddit OAuth authentication
• Caching layer (Redis) for embeddings and results
• Additional data sources (Product Hunt, Twitter/X, RSS feeds)
• User preferences (adjustable scoring weights)
• Search history
• Bookmarking/saving articles
• Export feeds (JSON, RSS)
• Real-time updates via WebSockets
• Advanced filtering options
• Dark/light mode toggle

ISC

Contributions are welcome! Please feel free to submit a Pull Request.

For issues and questions, please open an issue on the repository.

Built with ❤️ using Next.js, GraphQL, and OpenAI