Professional Overview
I'm a Data Professional specializing in Machine Learning, Applied AI, Data Science, and production ML systems. I help companies build production-ready solutions using Python, scikit-learn, XGBoost, LightGBM, PyTorch, FastAPI, MLflow, GCP, LangChain, and modern LLM tooling. My work includes fraud and anomaly detection, forecasting, model evaluation, probability calibration, explainable AI with SHAP, real-time inference, drift monitoring, and automated retraining. I’ve worked on ML systems handling 450K+ daily transactions, improved fraud-model ROC-AUC from 0.76 to 0.86, supported $35K+ in quarterly loss prevention, and built inference services delivering predictions in under 45 ms p95. I focus on building solutions that are accurate, deployable, measurable, and tied to a clear business outcome.
Experience
AI Expertise
Skills
Employment Type
Description of Responsibilities
- Evaluated ML outputs for risk, calibration, edge cases, and decision quality.
- Stress-tested models under missing data, drift, sparsity, and high dimensionalit
- Used SHAP and probability calibration to investigate model confidence and errors
- Built reproducible pytest checks for recurring model and evaluation failures.
- Validated RLHF evaluation batches using Python, Kafka, Redis, and deterministic
Key Achievements / KPIs
- Kept ~100K-row RL evaluation batches flowing through Kafka and Redis.
- Achieved zero acceptance failures through deterministic validation checks.
- Converted recurring model failures into reproducible automated pytest test cases
- Corrected an inflated high-risk recommendation using production base-rate calibr
AI Expertise
Skills
Employment Type
Description of Responsibilities
- Built production ML systems for fraud, forecasting, and real-time decisioning.
- Developed FastAPI inference services using Python, GCP, Pub/Sub, and MLflow.
- Built feature pipelines with SQL, BigQuery, dbt, Airflow, and Python.
- Evaluated models using SHAP, calibration, drift tests, and production metrics
- Supported deployment, monitoring, retraining, canary releases, and rollback.
Key Achievements / KPIs
- Improved fraud ROC-AUC from 0.76 to 0.86 with under 2.5% false positives.
- Helped prevent $35K+ in quarterly losses from fraud and pricing abuse.
- Achieved under 45 ms p95 latency for real-time transaction scoring.
- Reduced forecasting RMSE by 32% and off-peak labor costs by 14%.
Education
Field of Study: Exploratory Data Analysis, Machine Learning, Neural Networks, Tuning Models
Field of Study: Software Engineering, Web Development, Data Structure
Areas of Expertise
Data Science & Analytics
Extract business insights from raw data using statistical modelling, predictive analytics, A/B testing, and interactive dashboards.
AI & Machine Learning Engineers
Design, train, and ship production ML models — from data pipelines to model inference at scale. Covers supervised learning, neural networks, LLM integration, and model optimisation.
