A Streamlit-based intelligent exploratory data analysis application for uploaded CSV datasets.
The Problem
Initial dataset exploration often requires repetitive manual work such as profiling columns, checking missing values, understanding distributions, detecting outliers, reviewing correlations, and deciding whether a dataset is ready for machine learning.
The Solution
A guided analytics application that performs these steps automatically and organizes them into a structured analysis workflow.
Key Features
CSV upload
Dataset preview
Automatic data-type detection
Missing-value analysis
Duplicate detection
Descriptive statistics
Interactive Plotly visualizations
Correlation analysis
Outlier detection
Data-quality scoring
ML-readiness assessment
Target suggestions
ML algorithm recommendations
Business/data insights
PDF report generation
Technology / Architecture
Python
Streamlit
Pandas
Plotly
Scikit-learn
ReportLab
1
CSV upload
2
Profiling workflow
3
EDA and visualization modules
4
ML-readiness checks
5
PDF report generation
Challenges & Learnings
Challenges
Organizing many exploratory checks into a workflow that remains readable for different CSV datasets.
Presenting automated recommendations without overstating what the data can support.
Learnings
Automated EDA is most useful when results are structured, inspectable, and paired with clear caveats.
Report generation benefits from concise summaries instead of raw chart dumps.
