Hire Khaja A. - MLOps Engineer

Built and scaled end-to-end MLOps pipelines for machine learning models (fraud detection, credit risk), reducing deployment time from 8 to 4 weeks using MLflow, Docker, Kubernetes, and AWS. Developed AI-powered real-time decision systems for credit scoring and fraud detection, handling 5,000+ requests/min with ~99% availability. Applied modern ML techniques (TensorFlow, PyTorch, scikit-learn) to deliver personalized credit recommendations, increasing user engagement by 30%+. Optimized ML infrastructure and lifecycle management, reducing cloud costs by 20% through automation and efficient resource utilization. Improved model performance and reliability with continuous monitoring, validation, and governance, boosting fraud detection accuracy by 22% in a regulated environment.

Developed and deployed RESTful APIs using Flask, enabling scalable backend services and seamless system integration. Built data analytics pipelines using Python (Pandas, NumPy), improving data-driven decision-making by 40%. Integrated third-party APIs and designed real-time data processing workflows, increasing system efficiency by 23%. Optimized database performance using SQL, SQLAlchemy, and query tuning, reducing execution time by 30%. Implemented data validation, data quality checks, and ETL processes to ensure high data integrity and reliability. Collaborated with data science teams to integrate and deploy machine learning models, improving predictive accuracy by 20%.

Deployed scalable machine learning models using AWS SageMaker, reducing model deployment time by 25%. Built and optimized predictive and risk assessment models using statistical analysis, improving system quality and reducing defects by 15%. Designed and maintained CI/CD pipelines with Docker, Kubernetes, and Jenkins, reducing deployment failures by 35% and release cycles by 40%. Developed anomaly detection models for quality assurance, achieving 95% accuracy in identifying system risks. Created reusable ML components and automation workflows, improving team productivity by 40% Conducted A/B testing and model evaluation to enhance performance, increasing prediction accuracy by 10%.