TRACITY

• 

  #TRACITY

Hi everyone! For the MongoDB Challenge, we were asked to pick any public dataset and use AI, MongoDB’s powerful search and vector capabilities, and Google Cloud tools to create a solution that helps users better understand the world through data.

🔍 The Problem There’s plenty of data out there—on crime, air quality, COVID, energy use—but most people can't interact with it easily. It's either too scattered, too complex, or not user-friendly.

💡 Our Solution So, we built an AI-powered chatbot assistant that connects directly to real public datasets stored in MongoDB Atlas. Instead of users needing to browse tables or charts, they can simply ask questions like:

“What’s the AQI trend in Delhi?”

“Which state consumed the most electricity in 2022?”

And the chatbot fetches and explains the answer—without using OpenAI, relying purely on MongoDB’s semantic search and vector capabilities.

⚙️ How It Works We imported public datasets into MongoDB Atlas.

Used MongoDB’s full-text and vector search to make the data searchable by meaning, not just keywords.

The chatbot interprets questions, runs optimized queries, and delivers answers instantly through a clean, scrollable UI.

It’s modular, functional, and built to scale.

🌍 Impact This tool helps people—students, citizens, researchers—interact with important data effortlessly. We believe it brings real insight to the public by transforming raw data into useful answers.

Thank you!

Built With

• css
• github
• javascript
• mongodb
• python
• shell