An AI mentor that knows the farmer, the commodity, and the season.
Mothusi Mentor for Farmers is a voice-first AI companion built for agri-operators. It speaks the farmer's language, knows their crops and livestock, and gives advice that fits the current season, the local market, and the farmer's MGS state. It is not a chatbot. It is the working interface a farmer uses every day.
One of thirteen workflows in the Agriculture layer. All connected to the same growth record.
Why it exists, and how it operates.
A smallholder farmer in Phokeng asks Mothusi a question in Setswana about whether to sell broilers now or in two weeks. Mothusi knows their flock size, age, feed cost trajectory, and the current market price band. It answers with a recommendation, captures the conversation as a signal, and updates the farm growth record.
A maize farmer asks how to read a soil test result. Mothusi explains in the farmer's language, recommends an action, and offers a short training module if the farmer wants to learn more. The training is matched to the farmer's tier and prior knowledge.
Pre-trained per operator: when a farmer joins an agri-programme, Mothusi is pre-loaded with their farm profile, commodities, prior interactions, and the programme-specific evidence schema. The mentor knows them from session one.
How it works in the field.
- 01
Farmer asks in their own language
Voice-first, multi-language (Setswana, Sesotho, isiZulu, isiXhosa, Afrikaans, English, and more). Text input also supported.
Mothusi AI - 02
Mothusi reads the farm context
Commodity profile, current season, cycle stage, prior interactions, farm location, and tier-relevant signals are loaded.
Mothusi AI - 03
Personalised advice generated
Recommendation is specific to the farmer's commodity, season, location, and current operational reality. Not generic content.
Mothusi AI - 04
Training matched if needed
If the question reveals a knowledge gap, a matched training module is suggested. The module is short, applied, and producing an output.
Learning Mgmt - 05
Action captured as signal
The conversation, the recommendation, and any farmer-confirmed action are written to the growth record.
Growth record - 06
Follow-up scheduled
Mothusi checks back in: did the advice work? Did the training get applied? The follow-up itself is a signal.
Mothusi AI
What this workflow produces.
Each item below is a structured artefact written to the operator's growth record. Programme officers, funders, buyers, and the operator themselves can read and act on every one.
- Conversation log (farmer-anonymised by default)
- Personalised recommendations issued
- Applied training outputs (forecasts, plans, checklists)
- Follow-up outcomes (did the advice work?)
- Farm decision history with traceable reasoning
Different audiences. Same workflow output.
| Audience | Value delivered |
|---|---|
| Farmers | A daily companion that knows their farm, in their language, available outside business hours. |
| Extension officers | A multiplier on case-load. Mothusi handles the routine; officers handle the judgement calls. |
| Programmes | Granular visibility into where farmers are getting stuck across the cohort. |
| Funders | A behavioural signal for agri-lending: which farmers are engaging with their record, which are not. |
Deploy this workflow in your agri-programme.
Whether your programme runs farmer development, vet dispatch, offtake-backed lending, or processor support, this workflow plugs into your existing operations and feeds the same growth record as every other Mothusi capability.
From business support to measurable enterprise development. Across sectors, countries, and real operating environments.