Active Devices10of 10 registered
LIVE PORTFOLIO DEMO · DEMO MODE — SIMULATED IOT DATA
Real-Time Data Engineering Pipeline
A smart-building IoT simulation showing event ingestion, schema validation, transformation, anomaly detection, dead-letter handling and live operational monitoring.
Live: JavaScriptChart.jsBackend: Python + FastAPIPostgreSQLMQTTDocker
Simulator10 virtual rooms
ValidationSchema & quality checks
ProcessIn-memory event handling
StorageReading persistence
DetectionRules + rolling statistics
DashboardLive observability
Events Processed00 events/min
Active Alerts0No anomalies
Pipeline HealthPAUSEDDemo ready
Dead-Letter Events0Rejected safely
Data Freshness—latest accepted event
STREAM
°C
Temperature
STREAM
kW
Energy Usage
THROUGHPUT
EPS
Events per 10 seconds
DEVICE FLEET
10 online
Smart Building Rooms
INGESTION
WAITING
Live Event Stream
| Device | Temp | Humidity | Energy | Status |
|---|
ANOMALY DETECTION
0 alerts
Live Alert Feed
No anomalies detected
Use “Inject Anomaly” to test the detection path.
DATA QUALITY
0 rejected
Dead-Letter Queue
No rejected events
Invalid payloads are isolated instead of crashing the pipeline.
OBSERVABILITY
latest 10
Pipeline Activity
ENGINEERING BACKEND
Instant browser demo + full backend implementation
This GitHub Pages version runs the recruiter-facing simulation in the browser. The repository contains the completed Python/FastAPI implementation with PostgreSQL-ready persistence, MQTT/Mosquitto ingestion, a real transformation layer, three anomaly-detection approaches, Docker Compose and automated tests.
Live demo: client-side simulation for zero-login public access. Backend: fully implemented in the repository, but intentionally not hosted behind this GitHub Pages site.
IoT Simulator
→HTTP / MQTT
→Validation
→Worker
→PostgreSQL
→FastAPI
→Dashboard
FULL BACKEND IN THE REPOSITORY
More than the browser demo
The server-side implementation includes FastAPI ingestion, Pydantic validation, PostgreSQL-compatible SQLAlchemy models, MQTT/Mosquitto messaging, dead-letter persistence, transformation/enrichment, rule detection, rolling 3-sigma detection, Isolation Forest, analytics APIs, Docker Compose and automated tests.
WHAT THIS PROJECT DEMONSTRATES
Data engineering beyond a notebook
Streaming mindsetContinuous event generation and ingestion
Data qualityValidation and dead-letter isolation
ObservabilityHealth, freshness, throughput and alerts
Anomaly detectionThreshold and rolling statistical checks