Skip to content
AgriSensor

Now onboarding early-access farms

See water stress in the root zone before your crop does.

AgriSensor connects the soil probes and weather stations already on your farm, checks every reading, and uses AI to forecast root-zone moisture for each block. Each morning your team gets a clear irrigation plan and ranked crop-risk alerts.

Request early access Explore the platform
Works with
Existing probes
Forecast horizon
7 days
Decisions
For each block
Sample data
10 cm 27.5%
30 cm 29%
60 cm 33.8%

Block B1 · Maize

Tasselling (VT)

Action needed

VWC 30 cm

29%

Crop ET

7.1 mm/d

Rain, 5 d

0 mm

Recommendation

Irrigate 22 mm now to keep 30 cm above refill through silking.

Now — overdue 6 h

Hardware-agnostic

Reads soil probes, weather stations, and flow meters you already own, over LoRaWAN, cellular, or vendor APIs.

7-day root-zone forecast

Moisture is forecast at each probe depth, so you know when a block will reach refill before it happens.

Block-level irrigation plans

Recommended depth, run time, and start window for each block, ready to approve or adjust.

Crop-risk windows

Microclimate data is checked against disease and stress conditions specific to your crops and their growth stages.

The problem

Farms collect soil data. Few can turn it into decisions.

Soil probes are now affordable enough to put in every block. The bottleneck is turning thousands of readings into the right irrigation and crop-protection decisions every day, block by block, before conditions change.

01

Readings without decisions

A 4-depth probe logging every 15 minutes produces about 380 readings a day. Most farms check one graph on a vendor portal, if they check at all. The people starting pumps don't have time to interpret curves.

02

Calendar irrigation

Without a forecast, blocks run on fixed schedules. Some get water they don't need, which leaches nitrogen and raises pumping costs. Others cross the refill point during flowering or grain fill, when yield is most sensitive.

03

Silent sensor failures

Air gaps, flat batteries, and damaged cables produce readings that look plausible. A drifting probe can push a block's schedule in the wrong direction for weeks before anyone notices.

04

Knowledge that doesn't scale

An experienced agronomist can read a moisture trace and weather forecast and decide on a plan. Doing that for 40 blocks across three farms every morning isn't realistic, so most decisions end up being guesses.

The cost

Water, energy, and fertiliser spent where they aren't needed, and yield lost where they are.

Over-irrigation

Higher pumping costs, nitrogen leached below the roots, and more root disease in wet soil.

Under-irrigation

Stress at flowering, tasselling, or fruit set reduces yield, and watering later doesn't recover it.

Late disease response

Spraying after symptoms appear costs more and protects less than acting during the risk window.

The solution

One system from probe reading to field action.

AgriSensor sits between your sensor hardware and your field team. It checks the data, models what will happen in the root zone, and turns that into specific, reviewable actions. Agronomists set the thresholds, managers approve the plans, and supervisors carry them out.

  1. Readings arrive from existing probes and stations

    LoRaWAN, cellular, or vendor cloud. No hardware lock-in.

    1/4

  2. Faulty data is caught before it drives a decision

    Drift, spikes, gaps, and low batteries are flagged automatically.

    2/4

  3. Each block gets a 7-day moisture and risk forecast

    Based on crop stage, soil, weather forecast, and history.

    3/4

  4. Your team receives a plan they can approve

    Depth, run time, and timing, with the reasoning shown.

    4/4

Core features

Six modules that work as one workflow.

Each module handles one step from raw reading to field action. They share the same block map, crop calendar, and audit trail, so a probe fault in the Sensor Hub is excluded from the forecast, and an approved plan is reflected in the weekly report.

Ingest & validate

Workflow: Connect · Validate

Sensor Hub

Connects soil probes, weather stations, and flow meters through LoRaWAN network servers, MQTT, or vendor APIs. Every reading is unit-normalised, mapped to a block and depth, and checked for drift, spikes, and gaps.

You can trust the data a decision is based on, and you find out a probe needs attention before it costs you a schedule.

Predict

Workflow: Forecast

Root-Zone Forecast

Machine-learning models combine probe history, weather forecasts, crop stage, and soil type to forecast volumetric water content at each depth for the next 7 days, shown against each block's refill and full points.

You see stress coming days ahead, not after leaves curl.

Recommend

Workflow: Recommend

Irrigation Planner

Turns forecasts into a daily plan for each block: application depth, run time for that block's system, and the best start window. Managers approve, edit, or skip each item, and every change is logged.

Apply the right amount of water to each block, and keep a record of why.

Protect

Workflow: Recommend

Crop Risk Alerts

Monitors canopy humidity, leaf wetness, temperature, and soil moisture against crop-specific rules and models for disease windows, heat stress, and waterlogging. Alerts arrive ranked by severity with a recommended action.

Scout and spray when conditions call for it, not by the calendar.

Explain

Workflow: Act & learn

Field Advisor

Ask questions in plain language, like "why is B1 flagged?" or "summarise this week for the owner", and get answers grounded in your farm's readings, plans, and logs, with the sources listed.

Field teams and owners understand recommendations without needing to read the graphs.

Record

Workflow: Act & learn

Field Reports

Generates weekly reports covering water applied per block, plan versus actual, risk events, and sensor health, then exports them for owners, auditors, or water-use reporting.

Records that would take hours to assemble are produced from data you already collect.

How it works

From probe reading to field action in five steps.

Setup is a one-time mapping exercise. After that, the cycle runs automatically every time new readings and weather data arrive.

  1. 01

    Connect

    Link your probes, weather stations, and flow meters. Map each device to a block, crop, depth, and irrigation system.

    • LoRaWAN network server
    • MQTT / vendor API
    • Block & depth mapping
  2. 02

    Validate

    Incoming readings are cleaned, normalised, and screened for drift, spikes, gaps, and dead batteries.

    • Unit normalisation
    • Drift & spike checks
    • Gap filling rules
  3. 03

    Forecast

    Models forecast root-zone moisture and crop-risk conditions for each block over the next 7 days.

    • Crop stage & soil type
    • Weather forecast
    • Per-depth VWC model
  4. 04

    Recommend

    A daily irrigation plan and ranked alerts, each with its reasoning and supporting data.

    • Depth · run time · window
    • Ranked risk alerts
    • Reasoning attached
  5. 05

    Act & learn

    Approve, edit, or skip each item. Outcomes are fed back so recommendations improve for every block.

    • Approve · edit · skip
    • SMS / email to field
    • Outcome feedback

Interactive product tour

Explore a working AgriSensor workspace.

This is a fictional 212-hectare farm with eight blocks and eleven devices. Review alerts, approve irrigation, filter the sensor network, and ask the Field Advisor a question.

Sample data

Kestrel Ridge Farms

Good morning

Here's what needs attention across Kestrel Ridge today.

Root-zone moisture · Block B1

Maize · Tasselling (VT) · 14 days history, 7 day forecast

15%32%49%D-14D-7TodayD+6
10 cm 30 cm 60 cm Refill 34% Target 34–48% Forecast

AI insight · this morning

B1 needs water first. 30 cm is at 29% against a 34% refill point at tasselling. A 22 mm pivot pass restores it within about 12 h.

Hold C1. Deep moisture has been above full point for 3 days. Skipping one cycle saves about 110 m³.

Activity

  1. AgriSensor Raised a moisture-stress alert for B1 (maize, VT)

    08:14

  2. T. Mensah Approved irrigation plan for A2: 14 mm, Thu 06:00

    07:52

  3. AgriSensor Updated 7-day forecasts for 8 blocks using the 06:00 weather run

    07:30

  4. AgriSensor Opened a 48 h early-blight risk window for A1

    06:41

  5. R. Adeyemi Marked SP-1213 for a field check (C1)

    Yesterday

Sample data for demonstration. Kestrel Ridge Farms and its people are fictional.

Who it's for

Built for the people who decide when water goes on.

AgriSensor is designed for irrigated agriculture, where water is a managed input and timing affects yield.

Commercial growers

Irrigated vegetables, row crops, and orchards, 50–2,000 ha

Challenge
Water and energy are among the largest input costs, and a few days of stress at the wrong stage can cut yield.
How they use it
Daily irrigation plans for each block and alerts for disease windows.
Outcome
Fewer unnecessary irrigation cycles and fewer stress events at critical crop stages.

Farm & estate managers

Multi-site operations with field supervisors

Challenge
Managing many blocks on several farms, with decisions spread across WhatsApp messages and paper logs.
How they use it
A single view of every block and device, plus an approval queue and an activity history.
Outcome
Consistent decisions on every farm, with a record of who approved what.

Agronomy advisors

Independent consultants and input-company field teams

Challenge
Too many client farms to review sensor data by hand each week.
How they use it
A client portfolio view, cross-farm alerts, and reports to send to growers.
Outcome
Advice for more growers, based on data instead of site visits alone.

Cooperatives & outgrower programmes

Aggregators supporting many smaller irrigated farms

Challenge
Probes are deployed on member farms, but there's no capacity to turn the readings into guidance.
How they use it
Shared sensor networks, group alerts, and plain-language summaries for members.
Outcome
Members get guidance from the sensors on their farms, managed from one workspace.

Technology

How We Use Amazon Web Services

AgriSensor's architecture is designed around the path a soil reading takes. Devices connect through AWS IoT Core. Readings are validated by AWS Lambda and stored in Amazon S3 and Amazon DynamoDB. Forecasting models run on Amazon SageMaker AI, and Amazon Bedrock turns model outputs into plain-language advice. Results reach growers through secure APIs and SMS alerts. Each farm's data is kept separate, every device has its own identity, and the design scales with the number of probes rather than the number of servers.

  1. Stage 1

    Field Telemetry

    • AWS IoT Core

    Authenticates each device and receives probe and weather-station readings over MQTT.

  2. Stage 2

    Secure Storage

    • Amazon S3
    • Amazon DynamoDB

    Raw telemetry and reports stored in S3. Devices, latest readings, plans, and alerts in DynamoDB.

  3. Stage 3

    Processing

    • AWS Lambda

    Validates, normalises, and maps each reading to a block and depth, and flags sensor faults.

  4. Stage 4

    AI Intelligence

    • Amazon SageMaker AI
    • Amazon Bedrock

    Moisture and risk forecasting models, plus plain-language advisories grounded in farm data.

  5. Stage 5

    Application

    • Amazon API Gateway
    • Amazon Cognito
    • Amazon SNS

    Secure APIs for the dashboard, user sign-in for each farm, and SMS or email alerts for field teams.

AWS services in the AgriSensor architecture

  • AWS IoT Core: Connects soil probes, weather stations, and LoRaWAN gateways over MQTT, with a separate X.509 certificate for each device and rules that route readings to processing.
  • AWS Lambda: Validates and normalises incoming readings, detects sensor drift and gaps, and runs the daily irrigation-plan and alert jobs.
  • Amazon S3: Stores raw telemetry archives, model training datasets, and generated field reports.
  • Amazon DynamoDB: Holds the device registry, latest readings for each block and depth, irrigation plans, alerts, and approval history.
  • Amazon SageMaker AI: Trains and hosts the root-zone moisture forecasting and crop-risk models used for each block.
  • Amazon Bedrock: Generates plain-language field advisories and answers Field Advisor questions using the farm's own readings and plans as context.
  • Amazon API Gateway: Provides the authenticated API the dashboard uses, and receives early-access and contact form submissions from this website.
  • Amazon Cognito: Manages user sign-in and restricts each user's access to their organisation and farms.
  • Amazon SNS: Sends SMS and email alerts to field teams when a block needs irrigation or a risk window opens.
  • Amazon SES: Delivers website enquiry notifications to the AgriSensor team.
  • Amazon CloudWatch: Monitors device reporting rates, pipeline errors, and model job health, with alarms for missed data.

Early access

Bring AgriSensor to your farm this season.

AgriSensor was founded in 2026 and is onboarding a small group of irrigated farms and agronomy teams. Early-access partners work directly with the founding team, and their field feedback shapes the forecasting models and workflows.

  • Farm and block mapping

    We map your blocks, crops, irrigation systems, and existing devices with your team.

  • Sensor integration

    We connect the probes and weather stations you already run, or advise on where to add them.

  • Calibrated thresholds

    Refill and full points are set per block and crop stage, together with your agronomist.

  • Season review

    We review recommendations against what happened in the field and refine the models.

Request early access

Tell us about your operation. We'll reply by email.

Crops, hectares, irrigation systems, and any sensors you already use.

By submitting, you agree to our Privacy Policy.

Pricing

Plans that scale with your blocks and devices.

Pricing depends on the number of connected devices, blocks, and farms. Tell us about your operation and we'll send a quote.

Starter

Single farms getting started with sensor-based irrigation

Request pricing

  • Probes and weather stations for one farm
  • Root-zone forecast and irrigation plan
  • Sensor health monitoring
  • Email alerts
  • Weekly field report
Request pricing

Growth

Commercial operations with multiple blocks and field teams

Request pricing

  • Everything in Starter
  • Larger sensor networks across many blocks
  • Crop-risk alerts and approval workflow
  • SMS alerts for field teams
  • Field Advisor
  • Multi-user roles
Request pricing

Enterprise

Estates, advisory firms, and cooperatives with multiple farms

Request pricing

  • Multi-farm and client portfolio views
  • Everything in Growth
  • Custom crop models and thresholds
  • API access and data export
  • Onboarding for sensor networks
Contact sales

Team

The Team Behind AgriSensor

AgriSensor was founded in 2026 by Olu Gbenga to build software for irrigated agriculture. We work directly with early-access farms to set up their sensors, calibrate thresholds, and refine recommendations in the field.

  • Olu Gbenga

    Founder & CEO

    Leads AgriSensor's product direction, early-access farm partnerships, and company operations.

  • Joseph John

    Chief Technology Officer

    Leads AgriSensor's engineering, including the sensor data pipeline, forecasting models, and cloud architecture.

Talk to the founding team

Questions about pilots, sensor compatibility, or partnerships go straight to the founders.

olu@agrisensor.site Follow on LinkedIn

Contact

Talk to us about your fields.

Sales questions, sensor compatibility, partnerships, or press. Send a message and the team will reply by email.