Worked on the autoppia/autoppia_iwa and autoppia/autoppia_webs_demo repositories, delivering core backend features and stability improvements for local LLM services and web demo deployments. Focused on robust API integration, parallel LLM execution, and stress testing to validate performance under load. Enhanced deployment flexibility by enabling port-configurable multi-instance demos and improved reliability through dependency management and targeted bug fixes. Leveraged Python, Docker, and Bash to implement asynchronous processing, error handling, and observability enhancements. Expanded test coverage for large requests and edge cases, while maintaining clear documentation and configuration management to support scalable, resilient machine learning and agent-based system development.