The first step after loading your dataset.

Datapilot is an open-source Python library for deterministic exploratory data analysis and dataset understanding.

Every data project begins with understanding the data. Datapilot helps you understand your dataset before building machine learning models, dashboards, or AI-powered applications.

Install Datapilot from PyPI:

pip install datapilot-kit

import pandas as pd

from datapilot import analyze

df = pd.read_csv("dataset.csv")

report = analyze(df)

print(report.summary())
print(report.dataset_health())
print(report.insights())

Datapilot analyzes your dataset and returns a structured Report object containing dataset summaries, health assessments, statistical analysis, insights, and recommendations.

Datapilot currently supports:

• Pandas DataFrames
• CSV files
• Excel files
• Headerless .data files

• Dataset Summary
• Dataset Health Score
• Missing Value Analysis
• Duplicate Detection
• Data Type Analysis
• Statistical Summaries
• Outlier Detection
• Correlation Analysis
• Actionable Insights
• Professional HTML Reports

Project documentation is available in the docs/ directory.

• Vision
• User Journey
• API Philosophy
• API Review
• Roadmap

Datapilot v0.4.0 — Stable Release

Datapilot v0.4.0 extends the core data-quality workflow with:

• Dataset Health 2.0
• ML Readiness assessment
• ML assessment coverage
• Statistical Profile
• Data Integrity signals
• Notebook Readiness
• Assessment Boundaries
• Expanded HTML reporting
• Real-world dataset validation

The package is available on PyPI as datapilot-kit.

Datapilot is open source and released under the MIT License.