Salem, Oregon, United States
I build and evaluate AI systems that need to work reliably in real-world environments. My focus sits at the intersection of machine learning, data infrastructure, and system reliability, especially for applications in healthcare and high-stakes domains. Currently, I work on AI model evaluation and reasoning systems, where I analyze how models make decisions, identify failure modes, and improve alignment, transparency, and performance. I’ve built and supported data and ML pipelines using Python, PySpark, Airflow, Kafka, and AWS, and deployed systems using Docker and Kubernetes. I’m particularly interested in building scalable AI systems that are not just accurate, but trustworthy, auditable, and production-ready. I research trustworthy, secure, and scalable AI systems for healthcare, with a focus on real-world deployment and evaluation. My work sits at the intersection of three areas: healthcare AI (disease prediction, triage, clinical decision support), AI evaluation and reliability (probability calibration, uncertainty quantification, fairness, failure detection), and security-by-design (embedded security controls, adversarial robustness, governance). My differentiator is asking not just "does this model perform well?" but "is this system safe, calibrated, fair, and deployment-ready in practice?".
• Collaborate with AI researchers to evaluate and refine reasoning systems, ensuring transparency, integrity, and alignment of agent reasoning with defined machine learning objectives. • Analyze and annotate AI ideation traces to assess scientific honesty, validity, and compliance with established research standards through systematic validation of reasoning steps and model outputs. • Develop and modify Python-based machine learning workflows using scikit-learn and PyTorch, contributing executable code and analytical logic to advance experimental AI evaluation frameworks. • Engineered multi-trap benchmarks for senior backend engineering evaluation using a multi-step API audit across multiple heterogeneous input files, with many calibrated traps targeting cross-file synthesis failures and fabrication tendencies in frontier models.
During this period, I’ve been strategically investing in my personal and professional growth through a combination of advanced education, research, and practical project development. Highlights include: -- Doctoral Research in Artificial Intelligence and Cybersecurity Pursuing a Ph.D. in Information Technology at the University of the Cumberlands, focusing on applying AI and machine learning to healthcare cybersecurity, anomaly detection, and IoT security. -- Technical Upskilling Advanced technical expertise in AWS, Azure, GCP, dbt, PySpark, Airflow, Docker, Kubernetes, and MLflow through structured learning and real-world applications.
Contributed to the design and delivery of an advanced chatbot proof of concept (POC) using Python to streamline workflows, enhance accuracy, and improve user satisfaction. Applied Agile practices to foster collaboration and adaptability while optimizing data engineering processes for AI/ML systems. Key Accomplishments: • Automated AI/ML pipelines using PySpark, Airflow, Delta Lake, and dbt, streamlining ingestion, transformation, and model training processes, which improved development efficiency by 30%. • Built and demoed an AI-powered chatbot POC by integrating Python APIs, Hugging Face Transformers, LangChain, Dockerized microservices, and MLflow for tracking experiments and deployment. • Enhanced data infrastructure by implementing CI/CD workflows (GitHub Actions, Docker, Kubernetes) and leveraging Snowflake, Kafka, and AWS S3/Glue to ensure clean, scalable, and reliable pipelines across teams.