Fictional farm. Real management logic.
Agentic AI for farming, shown on a farm you can safely inspect.
Fallowbrook Farm is a clean-room teaching farm used to show what happens when livestock records, cash planning, documentation, and AI agents are put into the same operating system. The numbers are fictional. The logic is built to feel real to a working farmer.
North Yorkshire180 acresMixed livestock
“The bridge between AI capability and farm-level value is functionally free to build in money, not in time.”
Farmers
See how to build a working AI setup around real livestock, forage, and cash pressures.
Supply chain teams
Use Fallowbrook as a model for farmer-facing transparency, decision support, and skills uplift.
Public sector and levy bodies
Understand what scalable farm capability-building can look like in practice, not theory.
Layer 1
The video series
Ten chapters designed to move the viewer from curiosity to a working farm workflow.
01Free
End state
Why watch this series and what a farm operating with an AI agent actually looks like. The three core insights. The calf cost example as concrete proof.
02Free
Narrow AI vs generative AI
The distinction most agricultural AI coverage fails to make. Why the AI discussed at ag conferences is mostly narrow AI, while the AI that is actually transformational is generative AI.
03Free
The commodity of intelligence
The three main LLMs — ChatGPT, Claude, Gemini. What they cost, how they differ, where each is strongest. Practical demo comparing all three on the same farming question.
04Premium
Getting AI in the right place
Why the browser-based chatbot is limited. What the command line interface is and why it matters. The PTO shaft analogy — the CLI is the power take-off of your computer.
05Premium
Arms and legs
The jump from chatbot to agent. What changes when AI can read and write files, create folders, install tools, and interact with the filesystem.
06Premium
Copy as path
A specific, practical workflow shift. Instead of copying content into a chat, you give the AI the file path and let it read the file directly. Why this small change transforms quality.
07Premium
Dedicated folders and docs
How to structure a farm's digital workspace so the AI agent can navigate it. Naming conventions, folder hierarchy, maintained context documents.
08Premium
GitHub and Vercel
Version control and deployment as part of the farm operating system, not extra ceremony. Pushing a dashboard live so you can access it on your phone in the yard.
09Premium
Consistent conversation
Why the agent gets better when it lives around the work rather than only visiting occasionally. How exposure to context compounds over time.
10Premium
OpenClaw
The end-state vision. An open-source AI agent you interact with via WhatsApp, Telegram, or Signal from your phone. Messaging your farm's AI from the field.
Four layers of depth
From education to implementation
The site is designed to carry visitors deeper without forcing them to sign up or buy software first.
Layer 1Live scaffold
Video lessons
The public teaching surface. Ten chapters show the workflow shift from chatbot use to agentic farm operating.
Layer 2Live preview
Visible knowledge base
Visitors can see how the farm is organised: context, enterprises, financials, status notes, and the live app folders.
Layer 3Next build
API chat
A rate-limited public question box over the fictional farm repo so visitors can experience value without setup friction.
Layer 4Live modules
Apps and dashboards
Operational tools that prove the system can do more than answer questions: it can support decisions and daily rhythm.
Layer 2
The visible knowledge base
The point is not just to tell people that the farm is organised. It is to let them see what “organised” means.
fallowbrook-farm/ ├── README.md ├── CONTEXT.md ├── 01-business-overview.md ├── 02-enterprises/ ├── 03-financials/ ├── 04-metrics/ ├── 05-status/ ├── 06-analysis/ ├── 07-documents/ ├── 09-log/ ├── scripts/ ├── calf-dashboard/ ├── lamb-dashboard/ ├── cashflow-calendar/ ├── farm-control-tower/ ├── grazing-planner/ ├── mso/ └── site/
Core context
- README.md
- CONTEXT.md
- AGENTS.md
- MEMORY.md
Farm model
- 01-business-overview.md
- 02-enterprises/
- 03-financials/
- 05-status/
Reasoning and docs
- 06-analysis/
- 07-documents/
- 09-log/
- scripts/
Live app surfaces
- calf-dashboard/
- lamb-dashboard/
- cashflow-calendar/
- farm-control-tower/
- grazing-planner/
- mso/
- site/
Open the read-only structure browser
The public route uses a curated allowlist rather than exposing arbitrary repo contents.
Layer 3
Public API chat
This preview route is now wired for Anthropic and the same public-safe allowlist as the structure browser.
How many calves are active on the farm, and why does winter still constrain growth?
Fallowbrook currently has 86 active calves. Winter remains the binding constraint because the farm is designed backwards from housed forage pressure, not just summer grazing opportunity.
Open the Anthropic chat preview
The public chat is limited to the same curated allowlist used by the structure browser.
Layer 4
Working apps and dashboards
These are the portfolio pieces. They show what happens when the knowledge base is used to build real operating surfaces.
Calf dashboard
LiveTracks buying, selling, mortality, occupancy, and the economics of quick-turn dairy calf rearing.
Open appLamb dashboard
LiveShows the trading flock flow, draft pressure, margin logic, and seasonal use of summer grass.
Open appCashflow calendar
LiveMoves the conversation from livestock records into timing, liquidity, and the discipline of rent and loan cover.
Open appControl tower
LiveActs as the Monday cockpit for stock, cash, margin, winter pressure, and the next operational decisions.
Open appGrazing planner
LiveTurns the winter-pressure narrative into a seasonal grass and carrying-capacity view.
Open appDynamic MSO
LiveStress-tests stocking ambition against grass growth, forage carryover, and bought-in feed risk.
Open app