• 

  Home page of DataPilot AI

• 

  Interactive analytics dashboard

• 

  Dataset information and AI status

• 

  Data cleaning with missing value handling

• 

  Cleaned dataset preview

• 

  AI Analyst with statistical insights

• 

  Predictive Analytics model configuration

• 

  Machine learning performance metrics

• 

  Actual vs Predicted visualization

• 

  Dataset export and reporting

• 

  Downloadable prediction results and PDF report

Inspiration

Data analysis often requires using multiple tools for cleaning, visualization, machine learning, and reporting. I wanted to simplify this workflow by creating a single platform where users can upload a dataset and perform the complete analysis process in one place.

What it does

DataPilot AI is an AI-powered data analytics platform built with Python and Streamlit. It allows users to:

• Upload CSV and Excel datasets
• Automatically clean missing values and duplicate records
• Analyze datasets with interactive statistics and visualizations
• Train Machine Learning models using Linear Regression, Decision Tree, and Random Forest
• Compare actual vs predicted values
• Calculate performance metrics including R² Score, MAE, and RMSE
• Export cleaned datasets, prediction results, and PDF reports

How I built it

The project was developed using Python and Streamlit for the user interface.

Pandas was used for data processing and cleaning, Plotly for interactive visualizations, Scikit-learn for machine learning models, and ReportLab for generating downloadable PDF reports.

The application follows a modular architecture with separate components for data processing, analytics, machine learning, exporting, and reusable UI components.

Challenges I ran into

During development I faced several challenges, including handling datasets containing mixed data types, encoding categorical features, processing datetime columns, fixing model training errors, and generating professional PDF reports. Careful preprocessing and testing helped resolve these issues.

Accomplishments that I'm proud of

• Built a complete end-to-end analytics platform.
• Implemented multiple machine learning algorithms.
• Designed a clean dark-themed user interface.
• Added automated dataset health scoring.
• Generated downloadable CSV and PDF reports.
• Created a modular and reusable Python project structure.

What I learned

This project improved my understanding of data preprocessing, feature engineering, machine learning workflows, Streamlit application development, visualization, and project organization. I also learned how to prepare a complete software project for deployment and hackathon submission.

What's next for DataPilot AI

Future improvements include:

• AI dataset chat using Gemini API
• Time-series forecasting
• Automatic anomaly detection
• Model comparison dashboard
• Cloud deployment
• Advanced business intelligence dashboards

Built With

• git
• github
• numpy
• openpyxl
• pandas
• plotly
• python
• reportlab
• scikit-learn
• streamlit