ModMind

• 

  ModMind — Google Cloud Rapid Agent Hackathon Demo (MongoDB Track)

💡 Inspiration

Reddit moderators review hundreds of posts daily with no AI assistance. ModMind automates this with Google Gemini AI — giving instant toxicity scores, sentiment analysis, and moderation recommendations.

🛠️ What it does

• Fetches posts from any subreddit in real-time
• Analyzes content using Google Gemini via Vertex AI
• Returns sentiment, toxicity score, and mod action
• Stores every decision in MongoDB Atlas
• Shows subreddit statistics dashboard

🔨 How I built it

Built with React + Tailwind on Vercel, FastAPI on Render, Google Gemini AI via google-genai SDK, and MongoDB Atlas via MCP for storing mod decisions.

🚧 Challenges

• Gemini API quota management across free tier
• MongoDB URI encoding with special characters
• Wiring Vertex AI SDK authentication without Google Cloud billing account

📚 What I learned

How to build production AI agents with Google Cloud tools and integrate MongoDB as a real-time decision store for moderation workflows.

Built With

• fastapi
• google-gemini-ai
• google-genai
• mongodb-atlas
• python
• react
• render
• tailwind-css
• vercel