Ola Kowale — Data scientist, ML engineer, AI consultant

01 / About

Building things that learn

I am a data scientist and ML engineer based in Calgary who turns raw data into production-grade systems that drive real business decisions. My work spans customer analytics, statistical experimentation, and end-to-end machine learning pipelines across 13 different industry domains.

I build the full stack of a data science project: from exploratory analysis and feature engineering through model training and evaluation to deployment, monitoring, and documentation. Every one of my 23 public projects ships with reproducible code, clean notebooks, and honest performance metrics.

My approach favors rigor over hype. I care about experiment design, calibration curves, and whether a model actually outperforms a strong baseline. Good data science should be boring in its reliability and exciting only in the problems it solves.

Currently seeking data science roles focused on customer analytics, experimentation, and production ML systems

23

Projects

8+

Years Python

13

Domains

23

Public repos

02 / Skills

Tools of the trade

Machine learning and AI 95%

scikit-learn XGBoost LightGBM PyTorch TensorFlow LangChain Recommendation systems Predictive analytics

Data engineering 88%

SQL PySpark Airflow dbt Pandas Polars

Statistical analysis 92%

A/B testing Bayesian methods Causal inference Survival analysis Segmentation

Natural language processing 85%

Transformers spaCy RAG Embeddings FAISS

MLOps and deployment 82%

Docker MLflow FastAPI GitHub Actions AWS

Visualization 90%

Matplotlib Seaborn Plotly Streamlit Tableau Power BI D3.js

03 / Work

Featured projects

Anomaly detection

Fraud detection system

Real-time transaction scoring with ensemble methods, SMOTE handling, and threshold optimization for precision-recall tradeoffs.

View project

NLP / GenAI

RAG question answering

Retrieval-augmented generation pipeline over Calgary city documents with vector search, re-ranking, and answer grounding.

View project

Customer analytics

Churn prediction

Survival analysis and gradient boosting models predicting customer churn with SHAP explanations and retention strategy outputs.

View project

Experimentation

A/B testing framework

Bayesian and frequentist testing framework with power analysis, sequential testing, and automated decision rules.

View project

RecSys

Recommendation engine

Collaborative and content-based filtering with matrix factorization, cold-start handling, and online evaluation metrics.

View project

MLOps

MLOps pipeline

End-to-end ML lifecycle with experiment tracking, model registry, containerized serving, and automated retraining triggers.

View project

04 / Journey

How I got here

Education

MSc applied mathematics and statistics

Focused on stochastic processes, optimization, and statistical modeling. Built the mathematical foundations that inform every model I train today.

Milestone

Built 23 end-to-end ML projects

From problem framing through data collection, modeling, evaluation, and deployment. Each project includes full documentation and reproducible pipelines.

Milestone

Developed customer analytics pipelines

Built churn prediction, CLV estimation, segmentation, and propensity scoring systems that translate directly into actionable business strategies.

Milestone

Designed experimentation frameworks

Created A/B and multivariate testing systems with proper power analysis, sequential monitoring, and automated reporting for data-driven decision making.

Current

Exploring new opportunities in Calgary

Actively building production-quality projects and seeking a role where I can apply deep ML and analytics skills to real business problems.

05 / Contact

Open to opportunities

Let's build something together

I am always interested in hearing about data science and ML engineering roles, consulting projects, and interesting technical challenges.

Portfolio GitHub LinkedIn