Quantitative Risk Intern & System Architect
MTN Ghana
00. Impact
10,000
Monte Carlo simulations per scenario
87
Real-world scenarios stress-tested
18+
News feeds ingested every 15 minutes
90-day
Financial impact forecasting horizon
01. Overview
Led architecture and technical delivery for a four-person team building a real-time risk intelligence platform that replaced manual, subjective assessments with quantitative, traceable analysis.
02. The Company
MTN Ghana
MTN Ghana is the country's largest telecommunications operator, serving tens of millions of subscribers with mobile, data, and fintech services as part of the MTN Group — one of Africa's leading network operators.
mtn.com.gh03. Engineering Case Study
The Context
MTN was relying on manual, subjective assessments for financial risk. Leadership needed a quantitative system capable of forecasting exact financial exposure in Ghana Cedis under various stress scenarios.
The Architecture
I led a 4-person team to architect a real-time intelligence platform. I designed the ingestion pipeline to pull 18+ live news feeds every 15 minutes, feeding them into a 9-stage NLP and machine learning (XGBoost/LSTM) pipeline.
Technical Challenges
Providing a single risk number wasn't enough; leadership needed probability distributions. To achieve this, I engineered the backend to execute up to 10,000 Monte Carlo simulations per scenario without timeouts, accurately mapping out exact financial breaking points.
Measurable Impact
We successfully stress-tested 87 real-world scenarios and built a 90-day forecasting engine, completely shifting the enterprise from static, manual guesswork to real-time, automated financial forecasting.
04. Key Responsibilities
Led architecture and technical delivery for a four-person team building the real-time risk intelligence platform
Designed ingestion pipelines across 18+ news feeds refreshed every 15 minutes
Built a nine-stage AI/NLP pipeline that scored threats and estimated their financial impact in Ghana Cedis
Developed 90-day forecasting and reverse stress testing to identify the exact shock combinations behind financial breaking points
Enabled up to 10,000 Monte Carlo simulations per scenario, giving leadership a distribution of possible outcomes
05. Key Achievements
Architected a robust data pipeline that ingested and normalized data from 18+ live news feeds every 15 minutes.
Engineered a high-performance backend capable of running 10,000 Monte Carlo simulations per scenario to generate accurate probability distributions of financial risk.
Delivered a production-ready system that stress-tested 87 real-world scenarios, fundamentally transforming how MTN Ghana's leadership evaluated enterprise risk.
