LearnFarm is built for three kinds of people. Here is what each gets today.
Students, parents and curious professionals. Watch water and feed go in and eggs come out, day by day. One clear picture, instead of scattered videos and forums.
Thinking of starting a farm? Today, you can learn how the parts depend on each other. Later, a planner will take your city and budget and show what a flock, a greenhouse or a vertical farm would need and produce, with honest ranges and risks.
What if you add a crop, a greenhouse or a vertical unit? The goal is to model the change against published rates before you commit money. This part is not built yet.
Most of us eat three meals a day and have never seen how the food is made. LearnFarm shows you. You ask an AI questions in plain words, and a small 3D farm on your screen reacts as you learn. It is free, and every number comes with its source.
This is the real demo. Tap an area, open System, switch the well pump off, press play.
Forget tractors for a moment. Every part of a farm does two things. It takes something in. It gives something out.
Now the part most people miss. The output of one part is the input of another. Hen and cow manure goes to the compost heap. The compost feeds the tomato and corn fields. Waste from one place becomes food for the next.
Some things are stored: the water tank, the feed store, the compost heap. Stores can run out. When one does, the whole farm slows down. In LearnFarm, you watch that happen.
Once you see this, you can read any farm. Here is who it helps, and how you learn it.
Starting a farm is an expensive bet to make blind. LearnFarm lets you see how one works first, on a farm that costs nothing to break.
You talk. A small engine does the maths. The farm on screen shows the result. An AI coding agent is an assistant that can read and edit files on your computer. Here, it edits one file: your farm.
“Add 16 hens and make it rain.” Type it in plain English or Arabic, in Claude Code, Codex, Cursor or Gemini CLI.
The agent updates your farm file. Then a small open engine multiplies published rates, per hen or per square metre, by your numbers. The AI does not guess the totals.
Within a second, the screen updates. New hens walk in. Totals change, each with its source attached.
AI tools can sound sure and still be wrong. So LearnFarm does not let the AI invent a single rate. Every number traces back to a source you can open.
Kuwait is investing in growing more of its own food, in the desert, with a fraction of the water. LearnFarm helps more people understand how systems like these work.
large-scale indoor vertical farm in the Middle East opened in Kuwait: up to 550 kg of greens a day, using 90% less water.
US International Trade Administration ↗less water in the KFAS-funded vertical farm pilot in Wafra: 140 towers on 300 m² grew about 9,200 products in one month.
Kuwait Times, 2023 ↗Kuwaiti institutions worked with that pilot, including PAAF, KISR, Kuwait University and the Ministry of Education.
Kuwait Times, 2023 ↗Built in public, one phase at a time. Here is what works today, and what does not yet.
To look around: nothing. The live farm runs in your browser, on a phone or a computer. To talk to it: a computer with one AI coding agent set up, such as Claude Code, Codex, Cursor or Gemini CLI. Check each tool's own pricing.
No account, no build step. Copy the project, open the farm in your browser, then talk to your agent from the same folder.
# get it git clone https://github.com/BenjisCollector/learnfarm cd learnfarm # open the farm (http://localhost:5173/viewer/) python3 -m http.server 5173 # then, in your agent: # "add 16 hens and make it rain"