Meet Lima: AI-Powered Learning for Rural Agricultural Teams

How MOTHERLAND and peachstone.ai, supported by the Cisco Foundation, are building a WhatsApp-based internal learning system to enable Local Impact Teams to serve more farmer communities in Kenya by 2026.

The Challenge: Scaling Impact Without Losing Quality

At the beginning of 2025, MOTHERLAND found itself at a strategic turning point. Their Local Impact Team in Siaya County was performing exceptionally well—delivering regenerative agriculture training, conducting field demonstrations, and supporting farmer networks across several villages. The work was consistent, trusted, and deeply rooted in the community.

But it was also constrained by time, geography, and the physical realities of rural extension work.

MOTHERLAND's ambition to enter new communities was growing, driven by rising demand from farmers and local leaders. Yet the team's capacity couldn't expand at the same pace. Supporting more farmers didn't simply require "more staff"—it required a smarter internal system.

The need: A tool that makes local teams faster, more consistent, better coordinated, and less dependent on repeated in-person sessions.

This gap led to the birth of Lima.

Why Lima Exists: A Tool for Teams, Not Farmers

Unlike many digital agriculture tools that target farmers directly, Lima was intentionally designed for MOTHERLAND's Local Impact Team. Her purpose is not to replace trainers or digitize fieldwork—her purpose is to multiply their capacity.

Lima supports the team by:

Streamlining internal training processes
Organizing knowledge accessibly
Making learning materials available via WhatsApp
Ensuring training quality remains consistent across all villages

The challenges facing the team are significant. Climate shifts shorten planting windows. Soil conditions vary between neighboring farms. Market dynamics change weekly. Community training sessions often have to be repeated multiple times for different groups.

Lima acts as MOTHERLAND's extended internal digital backbone—a reliable, always-available learning companion that helps standardize work, guide new team members, and support expansion into new regions without diluting quality.

Built for Low-Connectivity Environments

Lima works entirely within WhatsApp because it's the most widely used communication channel across rural Kenya. This makes the system intuitive and instantly usable for all Local Impact Team members.

peachstone.ai developed infrastructure specifically adapted to MOTHERLAND's operating reality:

AWS-hosted backend supporting low-bandwidth environments
Semantic search engine for quick information retrieval
Structured onboarding and consent flow
Compressed PDF delivery and image optimization
Internal user-management system for tracking team interactions

The alpha version has been tested internally, successfully delivering learning materials, answering queries, and supporting the team during planning and preparation.

Why Lima Has a Personality

Even though Lima is used by staff and not farmers, she was intentionally designed to feel like a human companion rather than a technical system.

Research from the Stanford HCI Group and the work of Reeves & Nass in The Media Equation consistently shows that conversational agents with clear, relatable identities increase engagement, trust, and recall—even among professionals.

MOTHERLAND engaged women's groups, youth leaders, and community representatives to explore how people in Siaya communicate on WhatsApp. These insights shaped Lima's personality: warm, clear, friendly, approachable, and culturally relevant.

One participant summed it up: "She should sound like someone from here, like talking to a neighbor."

This philosophy guided her tone, rhythm, and way of explaining concepts. Even internally, Lima feels like a colleague—always available, always patient, always consistent.

Content Designed to Strengthen the Team

MOTHERLAND's content team created seven comprehensive learning modules:

Regenerative soil practices
Crop production
Composting
Erosion control
Post-harvest management
Market access
Safe digital behavior

These modules were transformed into compact, visually supported PDFs optimized for low connectivity.

For team members, the modules act as a unified reference system. Everyone trains with the same baseline knowledge, receives the same explanations, and prepares farmer-facing sessions with the same materials.

The hybrid model is powerful: Lima supports internal learning and preparation, while the Local Impact Team brings human connection, contextual knowledge, and practical demonstrations that only real field presence can deliver.

Real-Time Learning Loops

One of Lima's most important features is that she's not a static library—she's a dialogue partner.

Every time a team member asks a question, requests clarification, or opens a module, this interaction becomes part of a real-time insight system. MOTHERLAND can immediately see:

Which topics require more attention
Which areas cause confusion
Which modules need refinement

This continuous monitoring loop ensures MOTHERLAND grows based on evidence, not assumptions.

Built for Expansion

Every layer of Lima is engineered to support MOTHERLAND's expansion strategy. As the organization enters new communities, the Local Impact Team can rely on Lima to deliver foundational training internally within days, not weeks.

2026 Roadmap:

English and Kiswahili support
Structured internal learning journeys
Audio-based content
Step-by-step guidance for new staff

This capability allows MOTHERLAND to grow into new regions without compromising depth, accuracy, or consistency.

What Comes Next

The next development phase includes a structured pilot with the Local Impact Team to reach 50 more farmers. Their interactions will guide improvements in Lima's behavior, content flow, and training pathways.

In 2026, Lima is expected to evolve into one of the first WhatsApp-based internal agriculture companions in East Africa—designed to strengthen a field team rather than bypass it.

This approach reflects MOTHERLAND's philosophy: meaningful agricultural transformation requires trusted relationships, community presence, and human leadership. Technology should support the people doing the work, not replace them.

About the Partners

MOTHERLAND is a regenerative agriculture organization working with smallholder farmers in Kenya, focusing on sustainable practices and community-led development.

Cisco Foundation envisions a world of equitable, resilient, and empowered communities. Since 1997, it has partnered with organizations to create and scale innovative digital solutions that promote a healthy planet and advance the wellbeing of underserved communities globally.

Want to explore how AI-powered tools can support your organization's impact work? Get in touch.