Daniel Okafor
daniel.okafor@example.com+44 20 7946 0158linkedin.com/in/danielokaforgithub.com/danielokafor
Summary
ML engineer with 7 years building production recommendation, fraud, and forecasting systems.
Experienced across Python, PyTorch, feature stores, model serving, CI/CD for ML, monitoring, and distributed training on cloud infrastructure.
Experience
Senior ML Engineer
SignalCart | London, UK | February 2022 - Present
- Lead model-serving work for recommendation APIs handling 4,500 requests per second with 80 ms p95 latency.
- Built a feature store backed by Redis and BigQuery, reducing training-serving skew incidents from 7 per quarter to 1.
- Deployed PyTorch ranking models with shadow evaluation, canary releases, and automated rollback thresholds.
- Improved click-through rate by 8.6% through candidate-generation and re-ranking model changes.
- Mentor 5 engineers on ML design docs, offline evaluation, and production-readiness reviews.
Machine Learning Engineer
TrustPay | Cambridge, UK | June 2019 - January 2022
- Developed fraud detection pipelines using Python, Spark, Kafka, and XGBoost for 12M monthly transactions.
- Reduced false positive reviews by 23% while holding chargeback rate flat through threshold calibration.
- Automated model retraining and batch scoring in Airflow, cutting manual release work by 10 hours per week.
- Implemented drift dashboards for feature freshness, score distributions, and delayed label quality.
Data Engineer
RetailWorks | Nottingham, UK | September 2016 - May 2019
- Built ETL pipelines that processed 600GB of sales and inventory data per day.
- Implemented data quality checks that reduced failed forecasting jobs by 54%.
- Created Python feature pipelines for demand prediction used by merchandising teams.
- Optimised Spark jobs to cut nightly training-data build time from 4.5 hours to 95 minutes.
Skills
Languages
Python, SQL, Go, Bash
ML
PyTorch, TensorFlow, scikit-learn, XGBoost, Feature stores
Data
Spark, Kafka, BigQuery, Redis, Airflow
Platform
AWS, Kubernetes, Docker, MLflow, Prometheus
Education
- MSc Data Science, University of Edinburgh
- BSc Mathematics, University of Leeds