ML Universal Data Dashboard

A dashboard where anyone can upload any dataset, get it cleaned automatically, and receive AI-powered insights, patterns, and predictions — without knowing data science.

This is the fresh start (23rd Nov 2025).

Project Overview

This project aims to automate the most time-consuming parts of data preparation:

Loading datasets (CSV, Excel, JSON)

Detecting column types

Data profiling and summary reports

Cleaning and preprocessing

Encoding categorical columns

Preparing the data for machine learning models

The goal is to build a complete end-to-end data pipeline that works for any dataset with minimal manual setup.

Current Progress (Week 1)

Dataset loader

Column type detection

Data profiling module

Data cleaning module

Encoding module

Folder Structure project/ │ ├── src/ │ ├── loader.py │ ├── type_detection.py │ ├── profile.py │ ├── cleaner.py │ └── encoding.py │ ├── notebooks/ ├── docs/ │ ├── architecture.md │ └── problem_statement.md │ ├── data/ └── README.md

Roadmap

Add feature scaling

Add automated ML model selection

Add AI summary generation

Build the API layer

Build the frontend dashboard interface

Deploy as a usable web app

Tech Stack

Python

Pandas

Scikit-learn

Upcoming: FastAPI, Streamlit/React, Docker

Purpose

This is a long-term learning project where each part is built from scratch to understand the fundamentals clearly.