MyDataPilot

AI-Powered Data Science Assistant

What You Can Do With MyDataPilot

No coding or ML experience needed - just describe what you want

Work with Your Local Files

Just mention your file path - AI handles the rest:

  • "Analyze the data in /Users/me/sales.csv"
  • "Create charts from C:\data\report.xlsx"
  • "Combine all CSV files in ./data folder"

No uploading needed - files stay on your computer

See & Guide AI Thinking

Watch and modify how AI solves your problem:

  • View each step of AI's reasoning
  • Edit any decision or approach
  • AI adapts based on your changes

No "black box" - complete transparency

Get Expert-Level Results

AI writes professional code you can use:

  • Download Python scripts to reuse
  • Export publication-ready charts
  • Get detailed analysis reports

Learn by seeing how experts approach problems

Popular Tasks for Non-Technical Users

1
Sales Analysis

"Show me monthly sales trends from my Excel file and predict next quarter"

2
Customer Insights

"Find patterns in customer_data.csv and create segments"

3
Survey Analysis

"Summarize responses from survey_results.xlsx with charts"

4
Data Cleaning

"Clean messy_data.csv - fix errors and fill missing values"

Transform Your Questions Into Insights

Before MyDataPilot

  • ❌ "I need to learn scikit-learn first..."
  • ❌ "Which algorithm should I use?"
  • ❌ "How do I handle missing values?"
  • ❌ "What's cross-validation again?"
  • ❌ "My model accuracy is too low..."
  • ❌ "How to tune hyperparameters?"
With MyDataPilot

  • ✅ "Build a classifier for my dataset"
  • ✅ "Predict sales using historical data"
  • ✅ "Find customer segments in my CSV"
  • ✅ "Compare different ML models"
  • ✅ "Create a recommendation system"
  • ✅ "Detect anomalies in sensor data"

Real ML Example:

User asked: "Predict customer churn using my subscription data"

MyDataPilot delivered:

Data Preprocessing Feature Engineering Model Selection Cross-Validation ROC Curves Feature Importance

How It Works

AI-Powered Workflow

MyDataPilot Workflow: From user question to analysis results

From natural language question to professional data analysis - all automated with user control at every step

Watch MyDataPilot in Action

See a complete walkthrough of analyzing data with MyDataPilot

  1. Ask in Plain English
    Type what you want to analyze or create
    Example: "What are the key environmental factors that predict asthma rates across Illinois counties"
  2. AI Creates a Plan
    Review the steps AI will take to solve your problem
    Automatic decomposition into manageable analysis steps
  3. Watch It Work
    See each step execute with full transparency
    Edit analysis steps or code generation as needed
  4. Get Your Results
    Download code, visualizations, or insights
    Professional analysis results ready to use
MyDataPilot Interface

Main Chat Interface

Under the Hood

1
Natural Language Understanding

AI interprets your question and identifies the data analysis needed

2
Task Decomposition

Complex problems are broken down into manageable steps

3
Interactive Execution

You can edit and guide the AI at any step of the process

Why Choose MyDataPilot

No Coding Required

Just describe what you want in plain language. Our AI handles all the technical complexity behind the scenes.

You're Always in Control

Edit AI decisions, retry steps, or modify the approach. You guide the process while AI does the work.

See How AI Thinks

View and modify the AI's reasoning process. Understand exactly how your results are generated.

Your Data Stays Yours

Use your own OpenAI API key. Your data and analysis remain completely private and under your control.

Perfect For

Business Analysts

Create reports and dashboards without waiting for IT

Researchers

Analyze experimental data and create publication-ready plots

Students

Learn data science by seeing how experts approach problems

Entrepreneurs

Make data-driven decisions without hiring data scientists

Ready to Transform Your Data?

Join thousands of users who are getting insights faster with AI

No credit card required • Use your own OpenAI API key

How We Built It

Smart AI Architecture

MyDataPilot uses multiple specialized AI agents that work together to understand your request, plan the solution, write code, and verify results. Each step is transparent and editable.

AI Architecture

Built for Everyone

  • Natural Language Interface: No technical jargon required
  • Step-by-Step Execution: See exactly what's happening
  • Error Recovery: AI automatically fixes issues
  • Export Everything: Download code, data, or reports

Open Source: View our code, contribute features, or deploy your own instance.

View on GitHub

About MyDataPilot

Created at the University of Illinois Urbana-Champaign

MyDataPilot makes advanced data science accessible to everyone. Our mission is to empower professionals, researchers, and students with AI tools that amplify their capabilities without requiring programming expertise.

Xucheng Yu
Developer

Professor Haohan Wang
Project Advisor