AgroPulse
End-to-end greenhouse monitoring: the node reads temperature, humidity and soil, and drives the pump, fan, heater and door.
- Position
- Full-stack IoT developer
- Role
- Full build: firmware, backend and frontend.
- Period
- Mar 2026 – May 2026
- Status
- Complete

Context
IoT project for greenhouses joining an ESP32 field node, a cloud API and a web dashboard.
The challenge
Keeping greenhouse conditions stable means watching several variables at once and reacting in time; doing it by hand is slow and error-prone.
How I solved it
- ESP32-WROOM-32D firmware with PlatformIO: reads DHT11, DHT22 and soil moisture; drives pump, fan, heater and door through relays; shows status on an SSD1306 OLED.
- Readings sent as JSON over WiFi-HTTP, with automatic Serial USB fallback when offline.
- Java 17 + Spring Boot 3.2 REST API: Spring Security with JWT, anomaly detection, automation rules, SendGrid email alerts and a Resilience4j circuit breaker.
- Supabase (PostgreSQL) in production and SQLite in development; Docker container deployed on Render.
- React 18 + TypeScript dashboard with Recharts, MapLibre, shadcn/ui, GSAP and Lenis, deployed on Vercel using a Repository/Service pattern.
Outcome
- Complete system across three repositories: firmware, backend and frontend.
- GoF patterns (creational, structural and behavioral) and data structures in use: circular sensor buffer, irrigation queue and alert stack.
- Several iterative versions (v3, v5.1, v8) before the final form.
Contact
Let's talk. Write to me or browse my code.