AjoFlow is a hackathon-built digital savings-circle demo for Nigerian ajo/esusu groups. I developed the frontend, backend, and AI integration, with Esther Inyang supporting communications and Interswitch API testing and feedback.
The problem
Savings groups need a clear record of contributions, member reliability, and the order in which payouts happen. Cash and spreadsheet processes make those decisions difficult to inspect.
Hardest challenge
The payout queue had to respond to late and missed contributions without becoming a black box. The trust agent could suggest a new order, while admins needed to see the payment history and the reason for each change.
What I built
- Pod creation and contribution flows backed by Interswitch wallets, virtual-account callbacks, and disbursements
- A trust-scoring agent that evaluates payment history and explains changes to the payout queue
- A conversational AI assistant for navigating the product
- A Next.js frontend, Express API, and MongoDB data layer
- API tests for pod access controls and payout failure paths
Lessons learned
When AI influences a financial decision, the inputs and explanation need to stay visible so people can inspect the result. A score alone cannot explain a change to the payout order.
Evidence and status
Working demo including seeded accounts for exploring the dashboard and trust-scoring flow. The screenshot above uses that demo data.