AgroTutor is a shared advisory layer where partners plug in crop models, climate services and GenAI to deliver local advice over WhatsApp and other channels.
It sits inside the Digital Agricultural Network, connecting e‑Agrology data, farmer identities and advisory workflows so deployments can be reused across crops, regions and projects.
Local knowledge in one place
Bring together bulletins, trials and farmer feedback into a curated corpus for each use case.
GenAI‑ready advisory
Retrieval‑Augmented Generation turns that corpus into grounded, multilingual answers over WhatsApp and other channels.
Traceable interactions
Every question, answer and rating is logged and linked back to farms and plots for learning.
Part of a network
Works with e‑Agrology data and farmer digital identities to strengthen the innovation system.
Registry and routing layer where crop models, GenAI services and external APIs are plugged in and orchestrated per use case.
RAG pipeline, vector stores and prompt templates that turn local corpora into grounded advisory, with support for multiple languages.
Calendar, trigger and ticketing engine to send campaigns, route questions to experts and log every interaction across seasons.
AgMission weather maps and digital soil cards are model‑marketplace services that can be reused across channels and pilots.
This soil map and fertilizer card are a model‑marketplace service that can be reused across crops, regions and delivery channels.
Click on the map to select coordinates within the highlighted Bajío area, then choose parameter and crop to see the soil map and fertilizer recommendations.
Farmers ask questions via WhatsApp and AgroTutor routes them through GenAI and the model marketplace, with escalation to human experts when needed.
Farmer and plot registry
Link every interaction to a farmer, household, plot and season so advisory history is never lost.
Multi‑channel delivery
Deliver advice via WhatsApp, IVR, SMS or dashboards using the same underlying services and models.
Monitoring & governance
Logs, cost tracking and governance hooks so partners can comply with national data and AI guidelines.
Advisory scheduler
Schedule campaigns (e.g. sowing, fertilizer, pest alerts) and route consultations with expected response times and automated escalation rules.
Ticket lifecycle
Each advisory ticket has a unique ID, status and history so farmers see when advice is pending, answered or escalated, and managers can monitor service levels.
Feedback into models
Expert responses and farmer follow‑ups feed back into GenAI and rule‑based services as labeled examples to improve future advice.
The same infrastructure powers on‑demand chat and scheduled advisories driven by calendars, weather triggers or crop stages.
If you are exploring serious pilots or national-scale deployments and want to use AgroTutor as your advisory layer, write to us and we will respond with concrete next steps.
Please mention your country, crops and approximate farmer reach in the email.
Email us at s.nagaraji@cgiar.org