Automated Document Intelligence and Trade Risk Engine for SME Cross-Border Finance

Project

PFE-04 — Finance AI & NLP, Islamic FinTech, RegTech

Turn SME financial PDFs into financial ratios, a credit score, and a traceable Shariah compliance verdict in seconds.

Objectives

• Build document intelligence using LayoutLM and OCR to convert PDFs into clean JSON.
• Develop a Shariah rule engine producing traceable verdicts against a declared standard.
• Create a credit-risk model using XGBoost or LightGBM to generate a default-risk score.
• Build a FastAPI service with upload, review, and export capabilities.

Key challenge

Avoid silent misreads through confidence scoring and human-review routing.

Target

Evaluate extraction accuracy, Shariah agreement, and AUC against a ratio-only baseline.

Tech stack

Python, LightGBM, LayoutLM, FastAPI, PDFPlumber, Streamlit, XGBoost, Docker.

Profile

Computer Science or data science student at MSc or engineering level.

Format and location

• 1 intern
• Full-time, on-site internship
• 4 to 6 months
• Knowledge Place, Monastir, Tunisia
• Market: Malaysia → Gulf & Southeast Asia

General requirements

• Willingness to learn and take on the challenge
• Relevant skills and interests
• Adequate written and spoken English
• Ability to work alongside people from diverse backgrounds and disciplines
• Availability for long-term work after the internship