Crop-Mix Advisor
Spots the shortfall. Proposes the planting.
It reads the order book against what is actually in the racks, finds the gap, and tells you exactly what to plant to close it.
The most expensive mistake in indoor farming is planting the wrong thing. It does not announce itself. Six weeks later you are short on the crop the customer wanted and long on the one they did not, and both of them cost the same to grow.
The Crop-Mix Advisor nets committed demand — customer orders and standing weekly deliveries — against finished goods on hand and the cycles already in the racks. Where the two do not meet, it works backwards through the crop recipe: how many cycles of which crop, started when, arrive in time to close a 340-clamshell gap four weeks out.
It proposes. It never plants. The recommendation lands in the approvals inbox with its reasoning and its evidence attached, a human approves it, and the plantings that follow carry that person's name. Which is the difference between AI that helps you run a farm and AI you are quietly liable for.
Questions
The things people actually ask.
What does it actually look at?
The order book — one-off customer orders and standing weekly deliveries — finished goods on hand, and the cycles already growing. The shortfall is what you have committed to and are not on track to cover.
Can it plant on its own?
No. It files a proposal. A person with the right role approves it, and the resulting plantings are attributed to that person. There is no path from the advisor to the farm that does not go through a human.
How does it know how long a crop takes?
From the crop recipe. Stage durations are what let it work backwards from a delivery date to a planting date — which is also why the proposal is only ever as good as your recipes.
See Crop-Mix Advisor on a real farm.
We'll walk an order all the way through to its margin — on live farm data, not a slide.