Looking for a job

Kennedy Odhiambo

@kennedyodhiambo2

Data Scientist at Pesapal Limited, developing GNN models for fraud detection and AML to uncover complex financial crime patterns.

Kenya

Message

I am a Data Scientist and ML Engineer at Pesapal Limited, specializing in fraud detection, financial risk, and intelligent decision systems. I designed and deployed a machine learning pricing and profitability engine that leverages merchant and customer behaviour to recommend dynamic pricing rates, alongside a predictive credit risk model that identifies potential default risks early.

Previously, as a Junior Software Engineer/ML Engineer, I developed anomaly detection models for transactional streams that reduced fraudulent chargebacks by over 68%.

I also built end-to-end machine learning pipelines and containerized ML services using Docker, enabling reliable deployment of production ML solutions.My work includes graph-based merchant-customer analytics for uncovering complex fraud patterns, as well as NLP and speech-processing pipelines for analyzing customer interactions.

I hold a B.Sc. in Statistics from the University of Nairobi.

Experience

Work history, roles, and key accomplishments

Current

Data Scientist / Data Engineer

Current

Pesapal Limited

Jan 2026 - Present (9 months)

Designed and deployed a machine learning pricing and profitability engine using merchant and customer behavioural profiles. Engineered a predictive credit risk model and architected a centralized Business Intelligence ecosystem with Apache Superset.

Machine Learning Pricing Models Credit Risk Modeling PySpark Python Apache Superset ETL ELT Apache NiFi Data Pipelines SQL Docker Git Pytest API Integration

Junior Software Engineer/ML Engineer

Pesapal Limited

Feb 2025 - Dec 2025 (10 months)

Developed and deployed machine learning anomaly detection models in Python and SQL to monitor high-volume transactional streams, mitigating fraudulent chargebacks by over 68%. Implemented end-to-end ML pipelines and optimized high-concurrency SQL queries.

Python SQL Anomaly Detection Machine Learning Pytest Docker Data Validation Feature Extraction Performance Monitoring Query Optimization

Customer Experience Analyst

Pesapal Limited

Nov 2024 - Jan 2025 (2 months)

Formulated SQL models and cohort analyses to evaluate user interaction patterns and operational bottlenecks. Automated cross-departmental reporting pipelines, cutting weekly manual data extraction workloads and boosting operational efficiency.

SQL Cohort Analysis Reporting Data Analysis Automation