Chatlytics AI | Portfolio Project

Project Overview

Chatlytics AI is a production-ready conversation intelligence platform that converts exported WhatsApp chats into structured business intelligence. Instead of manually reading thousands of messages, users upload a WhatsApp TXT export and instantly receive executive dashboards, communication KPIs, behavioural analytics, Natural Language Processing outputs, Machine Learning insights and downloadable datasets.

The project demonstrates the complete analytics lifecycle: ingestion, parsing, cleaning, feature engineering, exploratory analysis, executive reporting and unsupervised machine learning inside a modern Streamlit application.

How to Export a WhatsApp Chat

Open the desired WhatsApp conversation.
Select More → Export Chat.
Choose Without Media for faster processing.
Save the generated .txt file.
Open Chatlytics AI and upload the exported text file.
The platform automatically parses and analyses the conversation.

Privacy: Processing occurs only on the uploaded export. No WhatsApp account access is required.

Data Processing Pipeline

WhatsApp TXT
      │
Regex Parsing
      │
Data Cleaning
      │
Feature Engineering
      │
NLP Processing
      │
Machine Learning
      │
Executive Intelligence
      │
Interactive Dashboard
      │
CSV Export

Parsing

Date extraction, user identification, multiline reconstruction and system-message detection.

Feature Engineering

Word count, character count, emoji count, media flags, URL detection, timestamps, temporal features and engagement metrics.

Visual Analytics

Interactive Plotly charts, KPIs, timelines, activity maps and executive summaries.

Analytics Engine

Total messages, words, links, media and emojis
Monthly and daily communication timelines
Most active participants
Word cloud generation
Top common words
Emoji frequency analysis
Conversation Health Score
Executive recommendations
Downloadable cleaned dataset

Natural Language Processing & Machine Learning

Chatlytics AI applies NLP preprocessing to transform unstructured conversations into analytical features. It then employs K-Means clustering to discover hidden communication patterns without requiring labelled training data.

Tokenisation
Word frequency extraction
Emoji and URL recognition
Behavioural feature engineering
Unsupervised K-Means clustering
Communication pattern discovery
Explainable executive summaries

Future roadmap: Sentiment analysis, topic modelling, transformer embeddings, conversational search, PDF reporting and predictive engagement forecasting.

Business Applications

Sales Teams
Measure customer engagement and communication efficiency.

Customer Support
Understand response behaviour and workload.

NGOs & M&E
Monitor programme communication and stakeholder engagement.

Research
Perform behavioural and communication analytics.

Healthcare
Study patient communication trends.

Education
Analyse student collaboration patterns.

Technology Stack

Python • Streamlit • Pandas • Plotly • Matplotlib • Scikit-learn • WordCloud • URLExtract • Emoji • Regular Expressions • Git • GitHub

Project Highlights

Transforms raw conversation data into strategic insights.
Integrates analytics, NLP and ML in a unified workflow.
Interactive executive dashboard designed for decision makers.
Modern Streamlit interface with downloadable outputs.
Demonstrates end-to-end data science engineering practices.

About the Developer

Amin Muhammed Badawi
Data Science Consultant specialising in Artificial Intelligence, Machine Learning, Business Intelligence and Monitoring & Evaluation Analytics.

Email: cigma.generalsolutions@gmail.com
Phone: 08065440075
LinkedIn: linkedin.com/in/elameenbadawy

💬 Chatlytics AI

AI-Powered Conversation Intelligence Platform