CASE STUDY / IOT · FULL-STACK
Smart Home ENMS
A residential energy project combining simulated device telemetry, consumption analytics, ML predictions, and bidirectional device control.
The system
A residential energy-management project connecting a React interface to FastAPI, PostgreSQL, MQTT, and Node-RED. Simulated smart-home devices publish telemetry at regular intervals, supporting consumption views, device controls, efficiency indicators, and cost scenarios.
Engineering decisions
- Kept telemetry ingestion separate from the API and interface so device messages and user requests have distinct responsibilities.
- Used MQTT and Node-RED for device communication and processing, with PostgreSQL for persistence.
- Exposed device state, energy analytics, and prediction features through a documented FastAPI interface.
- Containerized the application and supporting services with Docker Compose.
The Random Forest prediction feature belongs to an experimental, simulated-data workflow. No production accuracy, energy savings, or independently scored design claims are made here.
Platform Screenshots
NEXT PROJECT