Babs Olaniyi

Senior Data Scientist / AI-ML Engineer

Machine learning systems for healthcare AI, pricing, recommendations, and experimentation.

I am Babs Olaniyi, a data scientist at Causal Foundry in Barcelona. I build models, experiments, and data products that support healthcare financing, provider performance, pricing decisions, user engagement, and operational decision-making.

View projects Read CV LinkedIn

Current focus

Current work: healthcare AI, claims analytics, provider performance, capitation reform, anomaly detection, and decision-support systems at Causal Foundry.

7+ years across data science, ML, and analytics

9M+ insured individuals represented in healthcare analytics work

EUR11M+ annual pricing leakage protected through anomaly detection

Selected Work

Recent projects and research

Healthcare AI

Claims-driven health financing

Built models and pipelines for healthcare financing, provider performance, capitation reform, anomaly detection, and decision support across large-scale claims data.

Pricing and revenue

Pricing leakage and optimization

Developed anomaly detection, elasticity, uplift, and optimal-pricing models across a EUR500M+ product portfolio.

AI/ML engineering

Multimodal LLM recommendations

Built recommendation experiments that combine product reviews, metadata, images, temporal signals, and ranking metrics such as Recall, MRR, Hit Rate, and NDCG.

Open project

Writing

Latest posts

• Designing A Recommendation System For Search In Ecommerce Mar 2023
• Designing Machine Learning Solution For Course Recommendation Jul 2022

Publications

Research

• Data-Driven Approach to Capitation Reform in Rwanda2025
• Enhancing Product Recommendations with Multi-Modal LLMsISIR-eCom 2025
• User Engagement in Mobile Health ApplicationsKDD 2022

Reading

Recent favourite reads

Articles and papers I return to when thinking about recommendation systems, experimentation, ML systems, and practical data science.

Recommendations and ranking

Recommender systems

• Understanding Multimodal LLMs
• How LinkedIn Suggests People You May Know
• Adaptively Optimize Content Recommendation Using MAB Algorithms in E-commerce
• On the Usefulness of the Netflix Prize

Experiments and causality

Decision science

• Bandits for Recommender Systems
• Multi-Armed Bandits and the Stitch Fix Experimentation Platform
• Uplift Modeling in Python
• Conquering Cohort Analysis with Kaplan-Meier Estimates

ML systems and practice

Production data science

• Nobody Gets Promoted for Simplicity
• Debugging AI With Adversarial Validation
• Research Software Engineering in Python
• Data Distribution Shifts and Monitoring
• Introduction to Streaming for Data Scientists
• Building a Data Team at a Mid-stage Startup

Skills

Working toolkit

Python SQL PySpark Scala Google Cloud BigQuery Azure Databricks MLflow Airflow AWS PyTorch TensorFlow XGBoost LightGBM Scikit-learn A/B testing Forecasting Survival analysis Causal inference Recommender systems