An intelligence layer around every agricultural asset.
Zirat gives each pump, controller, machine and piece of infrastructure a persistent, evidence-backed record — an Asset Intelligence Profile — that accumulates identity, knowledge, evidence and history, and that people can ask questions of.
The asset is the central entity. Knowledge accumulates around it.
Farm systems organise around the season. Dealer systems organise around the sale or service ticket. Zirat organises around the machine — so the answer to “what do we know about this specific pump, right now?” does not depend on who is reachable by phone.
Seven layers, each adding context to the same record. Identity and Knowledge anchor the machine; Evidence and History record what has happened to it; Investigation and Recommendation reason over all of it; the Human Decision closes the loop.
What each layer represents, connects and contributes.
Asset Identity
Which exact machine this is.
- What it represents
- A permanent link between the physical machine and its record — the specific pump, not “a pump like it”.
- What it connects
- Make, model, variant and serial number
- Asset Anchor (QR or durable tag) or nameplate photo
- Owner, site, location and criticality
- Why it matters
- Knowledge without identity is generic. Everything else attaches to the right machine only if the machine is identified first.
- How it adds to the context
- Gives every document, photograph and service event one address to accumulate at.
Technical Knowledge
What is known about this make and model.
- What it represents
- The manuals, specifications and guides that describe how this equipment is meant to be installed, operated and maintained.
- What it connects
- OEM manuals and service guides
- Specification sheets and parts information
- Warranty terms and purchase records
- Why it matters
- A fault code or a seal type means little without the document that explains it — and the right revision of that document.
- How it adds to the context
- Supplies the general reference context, labelled as general equipment knowledge rather than fact about this unit.
Inspection Evidence
What was actually observed on this machine.
- What it represents
- Photographs, readings and notes captured at the machine, with who captured them and when.
- What it connects
- Photographs of the machine and its nameplate
- Operator and mechanic notes
- Readings and inspection observations
- Why it matters
- Evidence is the highest-trust layer: it is about this unit, not the category. It must never be blurred with inference.
- How it adds to the context
- Gives investigation something concrete to reason from, with provenance intact.
Asset History
What has happened to it over time.
- What it represents
- A persistent record of installation, inspections, services, repairs and replaced components.
- What it connects
- Installation and commissioning
- Service visits and repairs
- Replaced parts and ownership or site changes
- Why it matters
- When a mechanic or manager changes, the machine’s memory should not leave with them.
- How it adds to the context
- Lets a present observation be read against what happened before.
Investigation
Questions asked of this machine’s own record.
- What it represents
- Asset-specific questions answered from the machine’s connected knowledge, evidence and history.
- What it connects
- Plain-language questions
- Comparison of current observations with past records
- Missing-information flags
- Why it matters
- An answer is only as useful as its grounding. Scope is the machine’s own record, not the open internet.
- How it adds to the context
- Turns a folder of documents into something a person standing at the machine can ask.
Recommendation
Possible next steps, with sources.
- What it represents
- Source-grounded possible next checks, with observations and inferences kept visibly apart.
- What it connects
- Supporting records and documents
- Information gaps
- Safety and escalation guidance
- Why it matters
- A recommendation without its sources is a guess. Every claim should show what it rests on.
- How it adds to the context
- Shows what is known, what is missing and what a qualified person should check first.
Human Decision
The person decides. The outcome is recorded.
- What it represents
- The owner, operator or qualified mechanic decides; the confirmed action and its outcome are written back.
- What it connects
- Confirmed actions
- Recorded outcomes
- Audit trail of what was claimed, on what evidence
- Why it matters
- AI does not declare a machine safe or operational, and does not replace qualified inspection.
- How it adds to the context
- Folds each decision back into the history, so the next investigation starts better informed.
Answers grounded in the machine’s own evidence.
Zirat’s reasoning is scoped to one asset’s own record — not the open internet and not a model’s unguided recall. Every claim carries one of three labels, and the three are never blurred.
A documented fact about this specific machine, drawn from a logged record, photograph or note.
“This pump’s seal was replaced last year.”
Knowledge about the make and model as a category, from a manual or specification. Never presented as confirmed fact about this unit.
“The service guide specifies a seal inspection before extended running.”
A conclusion Zirat is reasoning towards from the available evidence. Labelled so, and always to be verified by a qualified person.
“The seal may need renewed inspection.”
Every investigation is structured, not a paragraph.
The specified answer format separates what is known from what is missing and what to check, and escalates rather than instructs on safety-relevant questions.
Zirat is decision support. It does not declare a machine safe or operational, and it does not replace qualified mechanical inspection.
- 01AssessmentWhat the evidence supports
- 02EvidenceThe records, photographs and documents it rests on
- 03Relevant historyEarlier events on this same machine
- 04Information gapsWhat is missing from the record
- 05Recommended first checksWhat a qualified person should look at first
- 06Safety and escalationWhen to stop and call a qualified provider
- 07Recommended next actionsPossible steps — for a person to decide
Designed to sit beside what a farm or dealer already uses.
Zirat is not a replacement for farm, dealer or maintenance systems. Where integration becomes valuable it is additive, at the edges. Integrations are planned, not live.
| System | Keeps owning | What Zirat adds |
|---|---|---|
| Farm management software | What was planted, applied and harvested | Knowledge about the equipment used to do the work |
| Dealer and service systems | Sales records, tickets and billing | A machine-level view of service history across providers |
| CMMS or ERP asset modules | Scheduled work orders and financial registers | The knowledge-and-reasoning layer about each asset |
| IoT and controllers | Telemetry and control | A place for sensor-derived facts to become evidence, later |
A deliberately narrow scope.
Zirat’s unit of intelligence is the physical agricultural asset — not the biology growing on the land it serves.
- A crop or plant intelligence platform
- An agronomy or yield-prediction tool
- A pest, disease or soil diagnosis system
- A weather intelligence platform
- A farm-management ERP
- A precision-agriculture telemetry platform
- A system that controls machinery autonomously
Specified, built on a shared foundation, not yet field-validated.
- Photo-based asset identification
- A shared asset knowledge foundation
- Evidence capture
- Asset history
- Grounded investigation
The foundation Zirat is designed to run on, live for other WeSimplifAI products.
- Agricultural equipment taxonomy
- Farm and dealer workflows
- Regional-language, low-bandwidth use
- Farmer and mechanic pilots
Specified in the Concept Note, Product Requirements and Product Specification. Not yet field-tested.
- ConceptCategory and boundary defined
- SpecificationProduct requirements and specification writtenWe are here
- Pilot designScope, partners and baselines agreed
- Field validationUsed by real operators on real equipment
- Measured outcomesResults published with methods and limits