Problem
Many Indonesian SMEs struggle to access formal financing because they do not have complete financial statements, even though they may have active sales and healthy daily transactions.
AI / Fintech Hackathon Prototype
Aksesa
AI-powered alternative credit scoring platform for Indonesian SMEs using OCR, ML, and LLM recommendations.
Role
Full-Stack Developer & AI Engineer
Year
2026
Stack
Next.js + FastAPI
Solution
Aksesa uses alternative business data, invoice OCR, machine learning scoring, and AI-generated recommendations to create a clearer creditworthiness profile for SMEs.
Overview
What I Built
Aksesa is a hackathon prototype for alternative credit scoring in Indonesian SMEs. It evaluates creditworthiness using alternative business data, invoice OCR, a machine learning scoring engine, and AI-generated recommendations to help non-bankable SMEs access fair financing opportunities.
Technical Highlights
Engineering Notes
REST API backend built with FastAPI
OCR flow for invoice/document extraction
Machine learning scoring pipeline with explainable factors
LLM-powered recommendation generation