Stefan Andelic Email & Phone Number

New York, Ny, Us

• Developed a cross-departmental, API-driven AI responsible governance "AI Readiness" application for collecting, transforming, analyzing, and interpreting large, complex datasets as well as performance & deployment metrics and assuring regulatory compliance across the complete lifecycle of an AI system.• Coordinated the integration of the IBM Chief Analytics Office's company-wide Python data science standards framework "Carrot" with the AI Readiness application.• Onboarded numerous applications (and growing) from across IBM onto the AI Readiness application.• Evaluated integration of third-party AI development tools and methodologies (e.g. HuggingFace etc.) within AI Readiness. • Fostered a collaborative environment that leveraged unique strengths of each team member and take advantage of effective communication & strategic planning to successfully tackle challenges posed by ongoing projects (e.g. WatsonX challenge project, Chatbot solutions etc.). • This application remains a cornerstone of the regulatory compliance strategy of the IBM Office of Privacy and Responsible Technology.

New York City Metropolitan Area

• Developed Identity Access Management (IAM) integration solutions for onboarding new and updating existing SaaS and IaaS IBM Cloud instances within Saviynt SaaS product.• Obtained the solution approvals through Architecture and Change Management boards. • Provided guidance to L1 support teams in implementing new Identity Access Management solutions and maintaining the existing ones. • Generated analytics reports with actionable features through SQL queries, runtime analytics control and custom queries.• Developed AccessHub Chatbot solution as an effective way for improving overall user satisfaction.

Completed projects include: • Designed, developed, and deployed a comprehensive conversational AI solution using IBM watsonx Assistant, incorporating intent classifiers, entity extractions, flow building, and integrating with Watson Discovery and watsonx.ai tools to enhance user self-service experiences. Successfully reduced costs by approximately $6 per query for fully contained conversations within the AI-powered solution.• Created and implemented a scheduling assistant solution powered by a Large Language Model, which streamlined user interactions across various platforms like Viber and WhatsApp. This innovation greatly improved appointment management efficiency and boosted user engagement and satisfaction. Additionally, it led to significant cost savings for service providers who are able to interface it from their Google Calendar accounts.• Successfully implemented a scalable, serverless ETL pipeline that streamlined data processing. The solution reduced operational costs by leveraging serverless architecture based on AWS S3, Lambda, Glue and Apache Airflow technologies, and provided timely insights for business decision-making. • Developed interactive and insightful Tableau dashboards to analyze and visualize performance, aiming to enhance business intelligence and support data-driven decision-making.• Successfully developed and launched a full stack application that significantly enhanced user engagement and operational efficiency. Achieved a 30% increase in user satisfaction through optimized performance and seamless user experience.