• Designed and deployed a production-grade pricing-date integration on GCP, for the Pharmaceutical Fee Schedule Calculator, supporting 300+ users with real-time pricing data.

• Developed and enhanced the Return-to-Work Supplement Program (RTWSP) application as a .XFDL form, improving frontend usability and reducing data entry errors by 30%+ for hundreds of external users.

• Built an internal search tool for RTWSP, connecting to Oracle SQL, allowing staff to query applicants by RTW#, first/last name, and quickly determine eligibility for supplemental checks.

• Integrated Google reCAPTCHA into public-facing applications to prevent automated submissions and enhance security across 100% of external traffic.

• Maintained operational reliability for the Qualified Medical Examiner (QME) online application, serving 200+ users with 99% uptime through monitoring and issue resolution.

• Led and managed a Project Portfolio with successful Cloud implementations at Fortune 500 companies.

• Developed and implemented numerous projects using Java, Python, and C++ to enhance functionality.

• Worked closely with development team members to identify and remove software bugs.

• Provided Cloud systems support for over 2000+ users.

• Participated in agile development of products within cross functional frameworks.

• Designed and developed scalable software within a scrum team.

• Conducted project planning, design, analysis, testing, and implementation of Cloud & Enterprise Resource

Planning systems.

• Performed root cause analysis of problems, documented faults in tracking system and generated daily reports.

• Maintained and supported Cloud Production environments, including programming, database maintenance, application security, and troubleshooting.

• Trained and evaluated large language models (LLMs) through diverse prompt-based tasks, improving chatbot output quality by 35% through rigorous annotation and review workflows.

• Curated and tested thousands of human-like responses, fine-tuning AI outputs for logical consistency, tone, factual accuracy, and empathy across various conversational contexts.

• Contributed to the development of robust, safe, and inclusive AI by identifying and flagging edge cases and bias patterns, reducing flagged harmful responses by 40%.

• Collaborated with an AI model training team to iteratively improve prompt templates and evaluation metrics, increasing overall model reliability scores by 25%.

• Maintained a >95% task quality rating across 500+ evaluations and 300+ hours of training data contributions used in model deployment readiness.

• Applied principles of natural language understanding (NLU) to design contextual prompts, enhancing model comprehension and response diversity in testing scenarios.