MicroCredit Advisor
Rural AI platform POC — two copilots on one shared platform
Personal project · independent demo
Independent demonstration themed on rural development — not commissioned, endorsed or deployed by IFAD. All borrowers, projects and figures are synthetic. The AI only advises; approvals are recorded by an authenticated person.
A working POC with two AI copilots: a Rural Finance Copilot for agricultural micro-credit, and a Project Monitoring & Impact Copilot for rural development projects — with cited answers and human approval.
Problem & business context
Loan officers in rural finance need to judge seasonal cash flow and risk from scattered documents, and development teams need to track project milestones, budgets and evidence — without giving decisions to an AI.
My role & contribution
Designed and built the POC.
What I built or designed
- Rural Finance Copilot: loan-officer workspace with seasonal cash flow, yield and price-shock scenarios, risk flags, answers cited from borrower documents, and human approval with a full audit trail
- Project Monitoring & Impact Copilot: milestones, budget (commitments vs. spending), indicators, missing reports, evidence-backed briefs, human report approval and a map of projects by country, region and city
- Farmer mobile app mock-up for sending crop photos and scanned documents
- Runs with Docker on a PC or AWS EC2; optional local Ollama model
