Digital Cultivation
One system. From the order to the margin.
The operating system for commercial indoor farming — production planning, closed-loop automation, per-plant vision, traceability, labor, and a real general ledger. One login. One source of truth.
The Problem
Most indoor farms don't fail on yield — they fail on operations.
The grow works. It's everything around it — the plan, the labor, the paperwork, the true cost of a pound — that's still living in spreadsheets and disconnected tools.
SOPs drift shift to shift
Steps get skipped, quality varies, and nobody can prove what happened.
Labor is guesswork
Manual scheduling burns hours and leaves operators unsure what's next.
Harvests miss the order
The grow plan isn't connected to what customers actually committed to buy.
Cost-per-pound is a black box
Without seed, labor, energy and packaging in one ledger, margin is a guess.
The Platform
One loop, from the order to the margin
Most farm software stops at monitoring. Digital Cultivation closes the whole business loop — and every step writes back into the next plan.
- 01
Order
Customer orders and standing weekly deliveries land in the order book.
- 02
Plan
CultivAid spots the shortfall and proposes the crop mix. The schedule back-calculates from the delivery date.
- 03
Grow
Recipes drive setpoints, lighting and fertigation. Robots and elevators work the racks.
- 04
See
Sensors validate conditions. Every plant is imaged and measured; outliers get flagged.
- 05
Harvest & Ship
Weights logged, packout run, shipment fulfilled — and the lot chain records all of it.
- 06
Margin
Cost-per-pound posts to the general ledger, benchmarked against the live market price.
…and the outcome feeds straight back into the next plan.
Modules
Everything an indoor farm actually runs on
Grouped exactly as they are inside the platform. Start with Farm Ops and switch the rest on as you grow — you're never paying for a module you don't run.
Every module has its own page — click any tile to go deeper.
Grow
The production core. Always included.Environment
Sense the farm, then act on it.Crop Intelligence
Every plant, measured.Quality & Compliance
Audit-ready by default.Business
Where the farm meets the P&L.Intelligence
AI that proposes. People who approve.A day with AGEYE
Less time chasing information. More time on the crop.
The platform organises the day around what actually needs a decision — and keeps the full operational record behind every one you make.
- 6:00 AM Brief
The morning digest is already waiting
CultivAid sends what changed overnight — cycles entering a new stage, alerts that fired, decisions waiting on you.
- 7:20 AM Triage
Work the worst rack first
Monitoring sorts by severity, so the rack that drifted is at the top — not buried in a wall of identical green tiles.
- 9:05 AM Approve
Sign off a setpoint the engine drafted
B-0513 entered grow phase 2. The recipe wants 17h of light; the schedule is still delivering 14h. Approve, and it commits with your name on it.
- 1:40 PM Inspect
Look at the plants that don't look right
This scan flagged 15 of 127 plants. You inspect those fifteen — not all one hundred and twenty-seven.
- 4:30 PM Close
Log the harvest
Weights in. Cost per pound out. The batch's COGS posts straight to the ledger before you leave.
- 6:30 AM Brief
See the whole floor before the shift starts
What's due, what's overdue, who's on, and what is blocked — across every room.
- 8:00 AM Plan
The day is already assigned
Seeding, transplanting, harvest and sanitation generated themselves from the production plan, by role and by shift.
- 11:15 AM Action
Resolve the exception
A task slipped, a zone drifted, plugs dropped below threshold. Route it, or approve the reorder CultivAid already drafted.
- 2:45 PM Verify
Check the work against the SOP
Sign-offs, photos and notes attach to the exact procedure version that was current when the work was done.
- 4:50 PM Close
Hand over the day
What's outstanding, what rolls to tomorrow, and what the next shift is inheriting.
- 7:00 AM Pulse
The whole operation on one screen
Active cycles, tasks, harvests due, approvals waiting, revenue month-to-date, labor hours this week.
- 9:30 AM Demand
Are we going to make the order?
Committed demand netted against finished goods on hand. The shortfall is a number on the screen, not a Friday meeting.
- 12:10 PM Approve
Approve the crop mix
CultivAid: plant four cycles of Genovese Basil to cover a 340-unit shortfall. Its reasoning and evidence are attached. You decide.
- 3:00 PM Margin
What did that pound actually cost?
Cost per pound on every harvested batch, benchmarked against the live USDA market price for that crop, that day.
- 5:30 PM Learn
Was the forecast right?
Model accuracy, per crop, cycle over cycle. Published back to you rather than quietly forgotten.
The plan writes the day — then runs it
Set a target — a thousand heads every Friday, or one delivery on a date — and the schedule back-calculates the plantings against your real constraints: mixed varieties, daily capacity, operating days, nursery and rack space.
Then it stops being a plan. Tasks generate themselves and land on the right person at the right hour. Recipes push setpoints to the lights, the pumps and the dosers. You choose how far the loop closes — advise, approve, or fully automatic — per system, not all at once.
Not a sample. Every plant.
AGEYE's imaging robots walk the racks and photograph every plant in the room. Each one gets measured — canopy area, height, greenness, temperature — and the ones that don't look like their neighbours are flagged for you automatically.
Canopy growth is tracked day over day, per tray and per plant, with time-lapse you can scrub. And because the platform knows what a plant should look like at that stage, yield forecasting stops being folklore.
The number every other farm platform can't give you
Seed, media, nutrients, packaging, labor hours and energy all land against the batch that consumed them. So when you harvest, you don't get a weight — you get a cost per pound.
Then it's benchmarked against the live USDA terminal-market price for that crop, on that day, and posted straight into a real general ledger. No re-keying. No shadow spreadsheet. No arguing about what a pound of basil actually cost to grow.
CultivAid
AI that proposes. People who approve.
CultivAid reads across every module — the schedule, the sensors, the scans, the labor log, the ledger — and answers in plain language. Then it does the thing nobody else will let an AI do safely: it acts.
It drafts. You decide. The audit trail remembers.
CultivAid never reaches into the farm on its own. It files a proposal — plant four more cycles of basil to cover a shortfall, extend the photoperiod for a cycle that just changed stage, reorder plugs that are about to run out — with its reasoning and its evidence attached.
Every proposal lands in one inbox. A human approves it. And when it executes, it executes with that person's name on it.
Built to be trusted
The boring things that decide the deal
Role-based access
Per facility, per section. Fail-closed by default — if a permission is missing, the door stays shut.
Human-approved automation
Setpoint changes and AI actions are proposals until a person signs them.
Fail-safe hardware
Actuators default OFF on stop or power loss. Safety is not a software promise.
Full audit trail
Who changed what, when, and what the system did about it.
FSMA 204 ready
Lot chain from seed to shipment, exportable, with mock-recall in a click.
API-first, ERP-agnostic
Push to your BI, accounting and ERP stack. No forced re-platforming.
Common questions
What growers ask first.
Every operation is different. A walkthrough starts from the systems, workflows and hardware you already have.
Do we have to replace the sensors and controllers we already have?
No. Third-party sensors and equipment connect through adapters, and the platform derives values like VPD and dew point at ingest regardless of who made the probe. Plenty of farms run Digital Cultivation on hardware we did not build.
Is this only for fully automated farms?
No. Every actuator class — lighting, dosing, climate — sits on its own autonomy ladder: monitor, advise, approve, or auto. You can run the entire platform with nothing in auto, and many farms do for a long time.
Can we start with one module instead of buying everything?
Yes. Farm Ops is the production core and is always included; every other module switches on when the operation actually needs it. You are never paying for a module you do not run.
Can different people on the team see different things?
Yes. Access is role-based per facility and per section, and it fails closed — if a permission is missing, the door stays shut. Growers do not see cost or margin surfaces unless you grant them, and CultivAid will not tell them either.
Does the AI change things on the farm by itself?
No. CultivAid can only file a proposal. A person with the right role has to approve it before anything executes, and the execution is recorded against that person in the audit trail — not against the AI.
Do we have to move our accounting into it?
No. The harvest COGS journal exports in a QuickBooks-ready format if your books live elsewhere. The value is that cost-per-pound is derived from operational reality instead of re-keyed from it.
Getting started
Start with Farm Ops. Switch the rest on as you grow.
You don't buy the whole platform on day one. Every farm starts with the production core — then turns on finance, or vision, or traceability, when the operation actually needs it.
Deploy with HYVE
Pre-integrated with AGEYE turnkey systems. Sensors, controls, imaging and automation connected out of the box.
Standalone SaaS
Run it on the farm you already have. Third-party sensors and equipment connect through the same APIs.
Connect your stack
Push to your BI, accounting and ERP tools. Bring your own hardware. Keep what works.
See it run a real farm.
We'll walk an order all the way through to its margin — the plan, the automation, the plants, the ledger — on live farm data.