HomePage
importing dataset
visualisation of data
View code source of the generated plot

DataLive.ai 🧠📊

🏆 1st Place Winner – LauzHack Mini Hackathon (GenAI Track) Organized by LauzHack and the EPFL AI Team

Inspiration

We’ve all dealt with messy datasets and repetitive code just to answer simple questions. We wanted a tool that would let anyone explore data just by asking — without writing code or digging through documentation. That’s how DataLive started.

Overview

DataLive.ai is a next-generation AI assistant for data exploration and visualization. It combines automated preprocessing, intelligent code generation, and multimodal reasoning — all through a sleek and interactive Streamlit app.

Upload a .csv file, clean your data, ask natural language questions, generate plots, and even receive AI-driven insights from those plots. All computation is performed locally, keeping your data private while maximizing speed.

This project was built in under 24 hours during the LauzHack Mini Hackathon 2025, where it won 1st place in the Generative AI Track.

Why It's Different from ChatGPT

Unlike general-purpose tools like ChatGPT or Code Interpreter:

Your data is never uploaded to the LLM → We extract only essential metadata (e.g., column names, types, number of missing values) and send that to the model. This:
- Preserves data privacy
- Avoids token overload on large datasets
- Ensures compatibility with sensitive or proprietary files
All code is executed locally → No need to copy/paste code; the app renders output instantly in the interface.
Lightning-fast interactions thanks to Streamlit → Near-instant feedback loop between code generation, execution, and visualization.

Features

Automatic Data Profiling

Data shape, summary statistics
Missing values report
Column data types and inferred structure

Smart Data Cleaning

Choose from:

Drop rows with missing values
Fill with column-wise mean
KNN Imputation

Conversational Code Generation

Powered by OpenAI GPT-4o
Ask natural questions like:

"Show a pairplot of numerical features" "Plot the class distribution as a bar chart"

Returns executable Python code, auto-rendered as:
- Seaborn heatmaps
- Histograms
- Boxplots
- And more...

Multimodal Plot Analysis

After generating a plot, ask:

"What can you conclude from this?"

The image is sent to Video-LLaMA2B via a custom Hugging Face inference handler
Returns textual interpretation of trends, outliers, and insights

Export & Traceability

Save all generated code
Export plots for reuse or reporting

Tech Stack

Frontend: Streamlit
Backend: Python (OpenAI API + Hugging Face Inference Endpoint)
LLMs:
- GPT-4o (OpenAI) – Prompt-to-code generation
- Video-LLaMA2B (DAMO-NLP-SG) – Plot interpretation
Libraries: pandas, NumPy, Scikit-learn, Seaborn, Matplotlib
Execution: Local sandboxed Python runtime (safe, fast, private)

How It Works

Upload your .csv dataset
Receive an automatic summary of key stats
Select a data cleaning strategy
Ask a question or request a visualization
Get back both:

The generated code
The rendered output
1. (Optional) Ask for AI-based interpretation of the plot

Getting Started

Launch Locally

streamlit run app.py

Environment Setup

You'll need:

OPENAI_API_KEY
HUGGINGFACE_TOKEN (for the Video-LLaMA endpoint)

Future Work

Here are some exciting features planned for the next iteration:

SQL Data Integration → Connect to relational databases (PostgreSQL, MySQL) and run natural language queries directly on live data.
NoSQL Support (MongoDB, Firebase, etc.) → Enable compatibility with document-based datasets.
Advanced Insight Generation → Incorporate models to automatically detect:
- Trends
- Seasonality
- Anomalies
- Correlations and causal patterns
Explainable AI Layer → Provide rationale behind detected trends or statistical recommendations.