A real-time delivery analytics system that simulates logistics platforms like Uber/DHL.
It processes streaming delivery events, applies ML-based delay prediction, and visualizes live metrics via a WebSocket-powered dashboard.
- Streaming: Kafka
- Processing: Python (Flink-style streaming)
- ML: Scikit-learn (delay prediction)
- Real-time layer: Redis + WebSockets
- Storage: PostgreSQL
- Frontend: Streamlit
- Deployment: Docker
- 🔥 Real-time Kafka event streaming
- ⚡ ML-based delay prediction
- 📡 Live WebSocket updates
- 📊 KPI dashboard (deliveries, delay, cities)
- 💾 PostgreSQL storage for analytics
- 🐳 Fully Dockerized system
git clone https://github.com/your-username/delivery-ops-dashboard.git
cd delivery-ops-dashboard
2. Create virtual environment:
python3 -m venv venv
source venv/bin/activate
3. Install dependencies:
pip install -r requirements.txt
4. Train ML Model:
python ml/train_model.py
5. Start Infrastructure:
docker-compose up --build
6. Create Kafka Topic:
docker exec -it kafka kafka-topics --create \
--topic delivery_events \
--bootstrap-server localhost:9092 \
--partitions 1 \
--replication-factor 1
7. Run Services:
PYTHONPATH=. python data_generator/kafka_producer.py
PYTHONPATH=. python streaming/flink_job.py
PYTHONPATH=. python realtime/websocket_server.py
streamlit run dashboard/app.py
8. Open Dashboard:
http://localhost:8501
🧠 What This Project Demonstrates
Real-time data engineering pipelines
Event-driven architecture
Streaming + ML integration
WebSocket-based live dashboards
Distributed system design
🔥 Future Improvements:
Apache Flink cluster (real)
React dashboard (frontend upgrade)
AWS deployment (MSK + RDS + EC2)
Grafana monitoring
Kubernetes (EKS)
👨💻 Author:
Subhankar Biswas