A powerful WhatsApp Chat Analyzer with beautiful visualizations
Live Demo • Features • Tech Stack • Installation • API
Chatlytics is a modern web application that transforms your WhatsApp chat exports into insightful analytics and beautiful visualizations. Upload your chat file and discover patterns in your conversations with 19+ analytics metrics.
Live Application: chatlytics.netlify.app
- Basic Stats — Total messages, words, and media shared
- Links Shared — Count of all URLs in conversations
- Most Active Users — Top 10 contributors (for group chats)
- Daily Timeline — Message activity over days
- Hourly Distribution — Peak hours of activity
- Weekly Activity — Day-of-week patterns
- Monthly Trends — Seasonal patterns
- Quarterly Growth — Long-term engagement
- Yearly Overview — Year-over-year comparison
- Initial Approach: Supervised Learning via TF-IDF Vectorization & Logistic Regression classifier.
- Current Engine: VADER (Valence Aware Dictionary and sEntiment Reasoner) for real-time rule-based intensity mapping.
- Hinglish Implementation: Augmented the VADER lexicon with a custom Hinglish/Roman-Hindi Lexicon (+300 words like zabardast, mast, badiya, bakwas) to handle code-switching in Indian chat contexts.
- Algorithm: LDA (Latent Dirichlet Allocation) from
scikit-learnfor unsupervised theme discovery. - Methodology: Uses Count Vectorization with a combined stopword engine (Standard English + Custom Hinglish Grammar) to extract semantic themes and their temporal distribution.
- Note on Old Chats: The Topic Evolution timeline focuses on the last 6 months of chat history. For older or sparse chats (fewer than 10 messages per month), the evolution chart may not appear if the data threshold is not met.
-
Mathematical Formula:
$$Score = (0.30 \cdot S) + (0.25 \cdot E) + (0.20 \cdot R) + (0.15 \cdot B) - P$$ -
Metrics:
-
$S$ (Sentiment): Positive/Negative compound ratio. -
$E$ (Engagement): Message frequency & active days. -
$R$ (Response): Temporal latency mapping. -
$B$ (Balance): Coefficient of Variation for participation. -
$P$ (Penalty): Anomaly deductions.
-
- Algorithm: Isolation Forest (Ensemble-based unsupervised outlier detection).
-
Impact Scoring: Statistical significance calculated via our custom Z-Score (
$\sigma$ ) formula:$$Z = \frac{\text{MessageCount} - \text{MeanCount}}{\text{StandardDeviation}}$$ - Dimensions: Multidimensional analysis of volume, sentiment, media bursts, and link density.
- Methodology: Behavioral mapping using multi-factor activity heatmaps and response latency patterns.
- Scope: Global Ranking Only (Reflects the intrinsic dynamics of the entire chat history, providing a static high-level behavioral overview).
- The 6 Behavioral Personas:
- Initiator — The conversation starter who kicks off new threads.
- Responder — The fast-action engine who keeps the momentum going.
- Driver — The primary volume steering force in the chat.
- Broadcaster — Detailed communicators with the highest word-to-message ratios.
- Listener — Concise participants who engage steadily with focused input.
- Night Owl — The late-night specialist who dominates after-dark activity.
- Per-User Analytics — Filter all stats by individual participant
- Dark Theme — Modern, eye-friendly dark UI
- Responsive Design — Works on desktop, tablet, and mobile
- Interactive Charts — Powered by Recharts with tooltips
| Technology | Purpose |
|---|---|
| Next.js 16 | React framework with App Router |
| React 19 | UI library |
| TypeScript | Type safety |
| Tailwind CSS 4 | Styling |
| Recharts | Charts and visualizations |
| Framer Motion | Animations |
| Lucide React | Icons |
| Technology | Purpose |
|---|---|
| FastAPI | High-performance Python API |
| Pandas | Data processing and analysis |
| Python Emoji | Emoji detection and counting |
| WordCloud | Stopwords filtering |
| Uvicorn | ASGI server |
- Node.js 18+
- Python 3.9+
- npm or yarn
git clone https://github.com/alexcj10/chatlytics.git
cd chatlytics# Install Python dependencies
pip install -r requirements.txt
# Run the backend server
uvicorn app.main:app --reload --port 8000# Navigate to frontend directory
cd frontend
# Install dependencies
npm install
# Run development server
npm run devThe frontend will be available at http://localhost:3000 and the backend at http://localhost:8000.
- Open WhatsApp on your phone
- Go to the chat you want to analyze
- Tap ⋮ (menu) → More → Export chat
- Choose Without media (recommended for faster processing)
- Save the
.txtfile - Upload to Chatlytics!
Upload a WhatsApp chat export file for analysis.
Request:
- Content-Type:
multipart/form-data - Body:
file(WhatsApp chat .txt export)
Response:
{
"users": ["Overall", "User1", "User2"],
"analytics": {
"Overall": {
"basic_stats": {
"Total Number of Messages": 5000,
"Total Number of Words": 25000,
"Total Number of Media Messages": 150
},
"links_shared": 45,
"most_active_users": {"User1": 2500, "User2": 2500},
"daily_timeline": [{"date": "2023-01-01", "message_count": 42}],
"hourly_activity": [{"hour": 14, "message_count": 120}],
"weekly_activity": [{"day_name": "Monday", "message_count": 850}],
"monthly_activity": [{"month": "January", "message_count": 1200}],
"quarterly_activity": [{"quarter": "2023Q1", "message_count": 3500}],
"yearly_activity": [{"year": 2023, "message_count": 15000}],
"most_busy_day": {"2023-12-25": 450},
"most_busy_weekday": "Sunday",
"most_busy_month": {"December": 4500},
"most_busy_hour": 21,
"response_time_analysis": {"User1": 45.5, "User2": 12.2},
"conversation_initiator": {"User1": 150, "User2": 140},
"longest_message": {"user": "User1", "message": "...", "date": "..."},
"most_wordy_message": {"user": "User2", "words": 150, "date": "..."},
"most_common_words": {"hello": 450, "thanks": 312},
"emoji_analysis": {"😂": 450, "❤️": 312},
"sentiment_analysis": {
"overall": "Positive",
"compound": 0.45,
"pos": 0.2, "neg": 0.05, "neu": 0.75
},
"user_sentiment_breakdown": {"User1": {...}, "User2": {...}},
"topic_modeling": [{"topic_id": 1, "words": ["vacation", "beach"]}],
"topic_timeline": [{"date": "2023-01-01", "topic_id": 1, "count": 5}],
"chat_health": {
"score": 82.5,
"rating": "Healthy",
"metrics": {"sentiment": 85, "engagement": 90, "response": 75, "balance": 80}
},
"anomalies": {
"spikes": [{"type": "Activity Burst", "date": "2023-12-25", "z_score": 7.08}],
"drops": [{"type": "Silent Period", "date": "2023-11-01", "gap_hours": 124}]
}
}
}
}The frontend is deployed on Netlify with automatic deployments from the main branch.
Live URL: chatlytics.netlify.app
The backend runs as a Docker container on Hugging Face Spaces.
# Uses Python 3.9 with FastAPI
# Exposes port 7860
CMD ["uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "7860"]chatlytics/
├── app/ # Backend (FastAPI Layer)
│ ├── main.py # API routes and orchestration
│ ├── analytics.py # Core statistical functions
│ ├── preprocess.py # WhatsApp chat parser
│ └── topics.py # Topic modeling orchestrator
├── ml/ # Machine Learning Layer
│ ├── anomalies.py # Isolation Forest (Outlier Detection)
│ ├── health.py # Chat Health scoring logic
│ ├── sentiment_vader.py # Enhanced Hinglish VADER engine
│ ├── sentiment_inference.py # Sentiment orchestration
│ └── topic_modeling.py # LDA-based theme discovery
├── frontend/ # Frontend (Next.js)
│ ├── src/
│ │ ├── app/ # App router pages
│ │ ├── components/ # React components
│ │ │ ├── Dashboard.tsx
│ │ │ ├── AdvancedAnalytics.tsx
│ │ │ ├── CRAPanel.tsx # Conversation Role Analysis Panel
│ │ │ ├── StatsCards.tsx
│ │ │ ├── UploadSection.tsx
│ │ │ └── Sidebar.tsx
│ │ └── lib/ # Utilities
│ └── public/ # Static assets
├── Dockerfile # HuggingFace deployment
├── requirements.txt # Python dependencies
└── README.md
This project is open source and available under the MIT License.
Contributions are welcome! Please feel free to submit a Pull Request.
- Fork the repository
- Create your feature branch (
git checkout -b feature/AmazingFeature) - Commit your changes (
git commit -m 'Add some AmazingFeature') - Push to the branch (
git push origin feature/AmazingFeature) - Open a Pull Request
Made with ❤️ by alexcj10