Journey Stories: Our Hackathon Experience

Our Action-Learning Journey

This is the story of how Team OSWEB navigated the CATS Hackathon journey, from early community visits to building AgriDatum, an AI-powered platform addressing agricultural data challenges in rural Nigeria.

Week-by-Week Reflections

Week 1 (11–17 November 2025): Discovery Phase

What we did:

During our first week, we visited farming communities in Akwa Ibom State. We spoke with farmers, market women, and NGO workers to understand the challenges of recording and verifying agricultural activities. We observed how farmers kept records and noticed the gaps in the documentation process.

Key learnings:

Farmers do not trust digital systems they do not understand. Data ownership and transparency are non-negotiable. Many farmers document their activities in their heads or on scattered paper. The lack of a unified, simple record-keeping system creates mistrust and prevents farmers from accessing support.

Challenges faced:

1. Farmers were busy with farming tasks and had limited time for long conversations
2. Some farmers struggled to articulate the problem because they saw it as “just how things are”

Highlights:

Hearing honest stories from farmers about how they keep records and why they do not trust external data systems. Realizing how much farmers want visibility but do not know where to start.

Week 2 (18–24 November 2025): Research & Insights

What we did:

We organized the interviews we collected and analyzed the stories we heard. We identified patterns across different conversations and grouped them into common themes like lack of reliable information, poor record-keeping, weak market linkages, and the absence of a unified community voice.

Key learnings:

Data is not the missing piece, structured, verifiable, and accessible data is. Farmers are willing to use tools that help them be seen and supported, not tools that exploit their information. NGOs and extension workers desperately need reliable data to improve resource allocation.

Challenges faced:

1. Keeping all the stories and insights organized without losing important details
2. Staying objective when forming insights and not jumping too quickly to solutions

Highlights:

Recognizing the recurring pattern that farmers and NGOs are both “blind” to each other’s needs. Realizing that farmers are willing to try new tools if the tools respect their work and offer immediate value.

Week 3 (25 November – 1 December 2025): Solution Design

What we did:

We brainstormed solution ideas based on our insights. We created the structure of AgriDatum, sketched out the “story hub” feature, and refined the core features. We evaluated multiple solution options and selected AgriDatum for its simplicity, feasibility, and multi-stakeholder benefit.

Key learnings:

Simple solutions with immediate value are more likely to be adopted than complex systems. An AI assistant can bridge the gap between data collection and actionable insights. The most important thing is trust. If the tool does not build trust with farmers, it will fail.

Challenges faced:

1. Deciding what to include in the MVP and what to leave for future phases
2. Balancing what farmers wanted with what we could realistically build in the hackathon timeframe

Highlights:

The moment we realized we could combine data collection with storytelling, making farming records both verifiable and personal. Understanding that AgriDatum is not just a database but a tool for unification.

Week 4-5 (2–13 December 2025): Building & Testing

What we did:

We built the first prototypes of AgriDatum. We tested the interface with farmers and adjusted the design based on their feedback. We integrated Cardano for data anchoring and added the AI farming assistant powered by Gemini API.

Key learnings:

Farmers respond positively when they see their records on-screen immediately. The AI assistant is a game changer. It gives farmers instant value while the platform builds their trust. Testing early and often prevents wasted effort building features users do not want.

Challenges faced:

1. Limited time to build everything we envisioned
2. Some farmers preferred verbal communication over typing, so we had to ensure the interface allowed voice-like simplicity

Highlights:

A farmer saying “If this thing can help us get buyers, we will use it every day” after seeing the harvest record screen. Successfully demonstrating the prototype to NGO partners and receiving immediate interest in using the platform.

Team Dynamics

What worked well:

1. Communication was smooth. We used WhatsApp and weekly check-ins to stay aligned.
2. Everyone brought unique strengths. Glory kept the process on track, Dominion led the technical build, Peace designed an interface that felt familiar to farmers, Favour ensured the backend and blockchain integration worked seamlessly, and Ola connected us directly to the community.
3. Feedback loop was effective. We listened to farmers, adjusted, tested again, and repeated.

What we’d improve:

1. We should have prototyped earlier in Week 2 instead of spending too much time organizing insights.
2. We could have divided tasks more evenly. Dominion and Favour carried a heavier load in the final weeks.

Roles & Contributions

Glory Archibong (Project Manager):

Led coordination, organized community research, managed timelines, and ensured we stayed aligned with the action-learning process.

Dominion Ekpuk (Frontend & Blockchain Lead):

Built the React-based interface, integrated Cardano using Aiken and BlockFrost, and implemented the farmer dashboard.

Favour Sunday (Backend & Blockchain Developer):

Built the Node.js backend, integrated PostgreSQL for off-chain indexing, and developed the API layer for the AI assistant.

Peace Essien (UI/UX Designer):

Designed the user flows, created wireframes, and ensured the interface reflected farmer preferences and cultural context.

Ola (Community Liaison):

Connected the team with farmers, organized field visits, and facilitated feedback sessions with the community.

Pivots & Changes

Major Pivots

Pivot 1: Data + Storytelling (Week 2)

• What changed: We originally planned to focus only on data recording. After listening to farmers, we realized they wanted a way to express their farming experiences, not just numbers.
• Why: Farmers said they wanted their work to be “seen” and “understood,” not just tracked.
• When: 20 November 2025
• Impact: AgriDatum became more than a record-keeping tool. It became a space where farmers feel respected and visible.

Pivot 2: Expanding to NGOs and Buyers (Week 3)

• What changed: We shifted from building only for farmers to building a platform that connects farmers with NGOs, buyers, and extension workers.
• Why: We realized the data farmers record only becomes valuable when it can be verified and used by support organizations.
• When: 28 November 2025
• Impact: The platform now serves multiple stakeholders, increasing its relevance and long-term sustainability.

Mentorship & Support

Mentors We Worked With

• Delfina – Helped us formulate insights from our ground truth research and stay grounded in evidence.
• Tabs – Provided feedback on simplifying the user interface and ensuring farmers could navigate the platform easily.
• NextTrend Hub – Guided us through the action-learning process and encouraged us to test assumptions early.

Resources That Helped

1. Cardano documentation for smart contract integration
2. Gemini API documentation for fine-tuning the AI assistant
3. Figma templates for designing mobile-friendly agricultural interfaces

Community Feedback

Feedback Sessions

Session 1: Prototype Testing with Farmers (6 December 2025)

• Participants: 4 farmers from Akwa Ibom (2 men, 2 women)
• Key feedback:
  • “The AI farming assistant is helpful, but the language should be simpler”
  • “We like that we can see our records immediately”
  • “Can we share these records with buyers directly?”
• Actions taken:
  • Simplified AI responses and reduced technical jargon
  • Added a “share” feature to allow farmers to download and send records via WhatsApp
  • Prioritized integration with market access tools in Phase 2

Breakthrough Moments

Our “Aha!” Moments

1. Storytelling is the connection. Realizing that farmers wanted to feel “seen,” not just counted.
2. NGOs need transparency, not just data. Understanding that the platform could help NGOs trust farmers and distribute resources more effectively.
3. First prototype screen was magic. Seeing a farmer’s face light up when they saw their harvest record displayed on the screen for the first time.

Challenges Overcome

Major Obstacles

Challenge 1: Building trust with farmers

• Problem: Farmers were hesitant to use a new digital tool without understanding who controlled their data.
• Solution: We explained that data is stored on Cardano, which means no single person controls it. We also demonstrated the “view and share” feature so farmers could see they retain ownership.
• Learning: Transparency and data ownership must be central to the platform, not an afterthought.

Challenge 2: Cardano integration complexity

• Problem: Integrating blockchain into the platform was more technically challenging than expected, especially with limited time.
• Solution: We focused on core functionality: recording harvest data on-chain and allowing farmers to view it. We postponed advanced smart contracts to Phase 2.
• Learning: Start small and build complexity incrementally. The MVP should solve the core problem, not demonstrate every possible feature.

Challenge 3: Limited farmer digital literacy

• Problem: Many farmers were not comfortable typing or using digital forms.
• Solution: We simplified the interface, used larger buttons, added voice prompts, and allowed Ola to assist with onboarding during field tests.
• Learning: Design for the lowest digital literacy level in your target audience. Simple, accessible design is a competitive advantage.

Personal Reflections

Individual Team Member Reflections

Glory Archibong:

“The CATS Hackathon taught me the power of listening. We did not come in with a solution. We came in with questions. That openness allowed us to build something farmers actually want. Working with this team showed me how much we can achieve when everyone brings their unique strengths to the table.”

Peace Essien:

“Designing for farmers taught me to see design as more than aesthetics. It is about respect. Every button, every color, every word on the screen must feel familiar and trustworthy to the people using it. I learned that good design honors the user’s world, not the designer’s preferences.”

Dominion Ekpuk:

“I learned the importance of early prototyping. We spent too much time planning and not enough time building in the first two weeks. Once we started testing prototypes with farmers, everything became clearer. Next time, I will prototype earlier and iterate faster.”

Favour Sunday:

“This hackathon reinforced the value of working closely with the community. Building for farmers is different from building for tech-savvy users. Simplicity, reliability, and trust matter more than advanced features. I also learned how to integrate blockchain into a real-world use case, which was challenging but rewarding.”

Ola:

“I realized how important it is to translate between the tech world and the community. Farmers speak in stories. Developers speak in features. My role was to help both sides understand each other. That bridge-building role taught me to value context, culture, and clear communication.”

Skills Developed

Technical Skills

1. Cardano blockchain integration (Aiken, BlockFrost, sha256, ed25519)
2. AI fine-tuning for context-specific use cases (Gemini API on African agricultural data)
3. Designing mobile-first interfaces for low-digital-literacy users

Soft Skills

1. Active listening and empathy-driven design
2. Community engagement and trust-building
3. Team collaboration and role clarity

What’s Next?

Post-Hackathon Plans

1. Onboard 50–100 pilot farmers in Akwa Ibom State
2. Partner with NGOs to test the dashboard and resource distribution features
3. Refine the AI assistant based on farmer feedback

Long-term Vision

AgriDatum will become the trusted platform for agricultural data in Nigeria and across Africa. Farmers will use it to record their work, access markets, and receive support. NGOs will use it to distribute resources transparently. Buyers will use it to find verified produce. The platform will create a new standard for transparency, trust, and empowerment in African agriculture.

Next Steps

View the complete technical implementation: Final Solution