AI Enabled Agricultural Advisory Solution Partner
Do you already run an AI advisory solution for farmers? Bring it to our benchmark and show us what it can do. The strongest solutions go on to the full request for proposals.
The challenge
Agriculture is the backbone of Rwanda’s economy. Twigire Muhinzi, our extension model, raises yields where it reaches farmers. But there are too few experts, advice is often generic, and farmers’ problems don’t reach decision makers in time.
Who the platform serves
Smallholder farmers
Mostly on feature phones and under half a hectare. They need direct answers by voice and SMS.
Extension officers and farmer promoters
The platform makes them faster and more consistent. It doesn’t replace them.
MINAGRI, RAB and agencies
They see what farmers ask, where the gaps are and how needs change, in real time.
What we’re looking for
A Kinyarwanda first advisory solution that works for crop and livestock advice, pest and disease management, and weather and climate guidance. It takes voice, text, images and USSD menus, and answers by voice, text or visuals over IVR, SMS, WhatsApp, apps and agent networks.
We prefer
- Conversational AI in Kinyarwanda
- Photo based pest and disease checks
- Designs that work on weak rural connections
- Open source models hosted in Rwanda
- Open standards and a modular design
We avoid
- Closed black box models that can’t be audited
- Lock in to one supplier for hosting, data or development
- Rwandan personal data stored abroad without NCSA approval
How selection works
-
Expression of interest
Firms send a profile, the technical checklist and a demo.
-
Benchmark (current)
We score each solution, share curated training data, then score it again.
-
Request for proposals
Top firms, alone or in a consortium, send a full technical and financial proposal.
Timeline
| Milestone | Date |
|---|---|
| EOI published | 28 September 2026 |
| Information session | 1 October 2026 |
| Submission deadline | 6 October 2026 |
| Demo Day | 7 and 8 October 2026 |
| Pre selection | 9 October 2026 |
| Onboarding | 15 October 2026 |
| Benchmark evaluation | October 2026 |
What to submit
- Organisation profile, ten slides at most, with your track record in AI farm advisory
- Technical checklist, completed and signed, from the annex
- Solution demo, a recorded link of 15 minutes at most, plus a live demo on Demo Day
How to submit
Email all items as PDF attachments, with the demo link in the email, to
procurement@c4ir.rw
Use this subject line
EOI for AI-Enabled Agricultural Advisory Solution Partner – [Company Name]
Technical checklist
Firms confirm each point. API access is required. For the rest, explain any gap and how you’d manage it.
- We give API access to our solution for the benchmark. Required.
- We can host in Rwanda and meet data residency rules under Law N° 058/2021.
- Our design is modular and takes new features without a rebuild.
- We share technical documents on our models, harness and data handling.
- We don’t lock C4IR, MINAGRI or RAB into one supplier.
- We don’t use a black box model that can’t be inspected.
- We don’t need special hardware that Rwandan teams can’t get.
What you’ll get to work with
| Dataset | Contents | Volume |
|---|---|---|
| Farm questions and answers | Questions from farmers with answers from agronomists | 5,000 pairs |
| English and Kinyarwanda sentences | Matched farm sentences to test translation | 2,000 sentences |
| Audio with transcripts | Farm audio for speech recognition and voice tests | 20 hours |
The selected partner also builds on a national agricultural corpus, published as a digital public good on the government data sharing platform.
This call is not a request for proposals or a binding offer. Only firms that take part in this EOI can bid at the RFP stage.