adnansyedzafer

Architected and initially developed an internal AI-powered QMS chatbot using Streamlit, later integrated it with existing .NET applications via FastAPI; leveraged LangChain, LangGraph, Ollama (Qwen 2.5), and MySQL to enable secure, real-time patient report assistance and automated billing insights. Developed multiple LLM-executed tools within the chatbot, including patient report status tracking, billing summaries, top-performing tests, and other on-demand queries, driven entirely by user input. Implemented contextual session handling, structured prompt governance, and role-based query scope, significantly reducing manual query resolution time and improving departmental efficiency. Designed and deployed interactive Power BI dashboards integrating cloud-hosted data from multiple departments, delivering real-time insights that supported strategic decision-making. These dashboards contributed to a 56% year-over-year revenue increase and optimized inventory management by reducing reorder frequency by 42%, enabling more accurate forecasting and resource planning. Integrated AI-driven automation and predictive analytics, enhancing data processing efficiency and reducing manual tasks. Assisted in the deployment, integration, and maintenance of scalable IT solutions, enhancing overall system performance, fault tolerance, and data management across enterprise applications. Delivered scalable IT solutions, training, and documentation, achieving 80% user adoption within one month and reducing support requests by 40%.

Played a pivotal role in integrating data science methodologies into software development processes, collaborating closely with cross-functional teams to streamline workflows and foster innovation in product development initiatives. Extracted and analyzed prospect and customer data from large databases, providing valuable insights for marketing and planning purposes. Implemented scalable and secure data management solutions. Created actionable insights through data visualization tools and analytics reports for internal stakeholders. Built machine learning models for prediction using customer behavioral data, shortening experimentation time. Assisted in software development processes alongside data science initiatives.