Skip to content

SETUP.md

Latest commit

 

History

History
67 lines (49 loc) · 2.09 KB

SETUP.md

File metadata and controls

67 lines (49 loc) · 2.09 KB
Raw

YC AI SUS Search Engine Setup

Fire trust (took me 2 hours to figure out Vercel 😭😭😭)

Environment Variables

Create a .env.local file in the root directory with the following variables:

# Pinecone API Key
PINECONE_API_KEY=your_pinecone_api_key_here

# Optional: Pinecone Environment (if needed)
PINECONE_ENVIRONMENT=your_pinecone_environment_here

# Firebase Configuration
FIREBASE_API_KEY=your_firebase_api_key
FIREBASE_AUTH_DOMAIN=your_project.firebaseapp.com
FIREBASE_PROJECT_ID=your_project_id
FIREBASE_STORAGE_BUCKET=your_project.appspot.com
FIREBASE_MESSAGING_SENDER_ID=your_messaging_sender_id
FIREBASE_APP_ID=your_app_id

Firebase Setup

  1. Create a new Firebase project at Firebase Console
  2. Enable Authentication and add Google as a sign-in provider
  3. Get your Firebase configuration from Project Settings > General > Your apps
  4. Add the configuration values to your .env.local file

Pinecone Index Setup

The application expects the following Pinecone indexes to exist:

  1. ai-agents - For AI Agents content
  2. looking-for-cofounder - For cofounder search content
  3. looking-for-internships - For internship opportunities
  4. side-projects - For side project content
  5. startup-advice - For startup advice content
  6. things-to-do-sf - For San Francisco activities

Vector Embedding

Important: The current API route includes a placeholder for vector embedding. You'll need to:

  1. Implement text-to-vector embedding (e.g., using OpenAI's text-embedding-ada-002)
  2. Replace the empty vector array in /src/app/api/search/route.ts with actual embeddings
  3. Ensure your Pinecone vectors are properly indexed with metadata

Running the Application

npm install
npm run dev

The application will be available at http://localhost:3000

Features

  • Glass morphic black search interface
  • Firebase Google authentication
  • Index selection for different content types
  • Real-time search with loading states
  • HTML-formatted results display
  • Responsive design for all devices
  • User profile display with sign-out functionality