Field notes, written by sensors

Your plants can't speak.
Your sensors can.

GrowSense listens to soil moisture, temperature, humidity and light, then turns the raw numbers into plain-language advice: what is wrong, how urgent it is, and exactly what to do next. Built for small farms, greenhouses and windowsill gardens.

Open the live dashboardNo signup. Three demo devices already growing.

Basil Shelf B2

Reading sensors

...

Soil

...

Temp

...

Humid

...

Lux

"Opening the greenhouse door..."

GrowSense insight, generated from live sensor data

How it works

Sensor to sentence in three steps

1. Listen

Any ESP32-class microcontroller posts a reading over HTTP: soil moisture, temperature, humidity, light. One POST per interval, that is the whole contract.

2. Understand

The agronomy engine compares every reading against crop-specific healthy bands and rolling trends, then writes a diagnosis in plain language with a severity level.

3. Act

Watch and critical findings surface as an alert ledger on your dashboard, so you fix the one bed that needs it instead of watering everything on a timer.

Live demo

A working greenhouse, not a mockup

The dashboard you are about to open runs on a real database with three simulated devices streaming readings every few seconds. One of them, Basil Shelf B2, is quietly drying out; watch the engine catch it. The same pipeline accepts your own hardware the moment you paste an API key.

• 48 hours of history per device, drawn as live charts
• Crop-specific healthy bands for tomato, basil, pepper and more
• Alert ledger that records every critical finding

Watch it think

Alert ledger

Basil Shelf B2 · live

Soil moisture drifting below the healthy band; the engine is watching it.

Tomato Bed A1 · live

All metrics inside healthy ranges.

Monstera Corner · live

Humidity creeping high; ventilation advised.

POST /api/ingest

curl -X POST https://growsense-ai.alikarki2580.workers.dev/api/ingest \
  -H "Content-Type: application/json" \
  -d '{
    "api_key": "gs_live_your_key",
    "soil_moisture": 54.2,
    "temperature": 23.1,
    "humidity": 61.0,
    "light": 480
  }'

For builders

Bring your own hardware

Add a device in the dashboard, get an API key, and point any microcontroller with a Wi-Fi chip at the ingest endpoint. An ESP32 with a capacitive moisture sensor costs about $6; if you have no hardware yet, the simulator shows you exactly what your data will look like. Every accepted reading returns the generated insight, so your device can even drive an on-site display.

Where it fits

Built for the real world

AI + Hardware Integration

An open ingest API for real sensors feeding an analysis engine that writes human-readable diagnoses.

Sustainability & Smart Cities

Water and crop savings for smallholders: treat the one bed that needs it, not the whole farm on a timer.

Smart Health Technology

The same band-and-trend engine generalizes to air quality and cold-chain monitoring.

Open Innovation

Every dashboard is also a public API: readings in, structured insights out.

Your plants can't speak.Your sensors can.