🧠 Project Capabilities Overview
- Dataset Upload & Management Users can upload their own CSV datasets through the frontend. Uploaded datasets are parsed locally for instant preview and analysis. The frontend sends the file to the backend, which: Concatenates all columns into a text field for each row. Generates vector embeddings for each row using Gemini. Stores the data and embeddings in MongoDB under a unique collection name. The backend returns the collection name, which is used for all future queries and analysis.
- Conversational AI Chat Users can ask natural language questions about the currently loaded dataset. The chat interface: Sends the user’s question and dataset ID to the backend. The backend performs a vector search in MongoDB to find the most relevant rows. Gemini generates a context-aware answer using the search results. The answer and top sources are displayed in the chat. If the backend is unavailable, the frontend falls back to local keyword-based search and provides basic insights.
- Automated Data Analysis Users can trigger automated analysis (e.g., “Analyze Data” button). The backend performs: Statistical summaries (row/column counts, unique values, missing data). Trend detection, correlation analysis, and other data science tasks. Returns results and suggested visualizations. The frontend displays these results in a modern, interactive panel.
- Data Visualization Users can request specific charts (histograms, box plots, bar charts, scatter plots, etc.). The backend generates chart data (using Plotly or similar) and returns it to the frontend. The frontend renders interactive charts for exploration.
- Preloaded Datasets The app includes built-in datasets (Olympics, GDELT, UN SDG) that can be loaded and analyzed instantly.
- Modern, Responsive UI Clean, modern React frontend with Tailwind CSS. Real-time chat, loading indicators, and user-friendly upload experience. Suggested queries and analysis prompts to help users explore their data.
- Extensible Backend FastAPI backend with endpoints for: Querying (vector search + LLM answer) Analysis Chart generation Custom dataset upload Easily extendable for new analysis types or data sources.
, AI-powered data exploration tool. Users can upload any CSV, chat with an AI about their data, get instant analysis and visualizations, and explore both built-in and custom datasets—all in a modern, interactive web app.
Log in or sign up for Devpost to join the conversation.