• Building industrial software that closes the gap between raw machine signals and shop-floor decision-making. Worked on an internal IoT middleware platform that transforms PLC data into contextualized operational insight for factory supervisors, using Python, TypeScript, and React.
• Developed an AI agent integrated with PC Factory MES to make production data immediately accessible through Telegram. Built with the Agno framework, it lets teams query live manufacturing information in natural language instead of navigating traditional MES workflows.
• Supporting delivery across multiple products by setting up CI/CD with Azure DevOps and Azure Pipelines, contributing to full-stack Node.js and TypeScript applications, and currently training computer vision models for defect detection in textile manufacturing.

Final-year Computer Science student focused on machine learning and software engineering. My undergraduate thesis explores mechanistic interpretability for coding large language models, studying how transformer-based systems internally represent and reason about source code.