Developer
CurrentDevelopment of Machine Learning applications, application of AI for internal efficiency and energy efficiency/performance optimization.- Currently working with hardware and Kubernetes optimization for high-performance applications, log anomaly detection, clustering, and classification.- Filed 3 patents relating to energy efficiency.- Developed a chatbot with a RAG system for internal semantic document search. Working with LLMs, prompt engineering, custom semantic search system capable of extending context window limitations.- Developed a novel generic AI-based approach for energy and performance optimization. Applied on optimizing network configurations, hardware settings including CPU core and fabric frequency scaling and C-states, and kubernetes (k8s) microservice settings.- Log anomaly detection and classification of large collection of logs. Developed a way to classify log collections with high accuracy using polluted data (99% contaminated, needle in a haystack scenario).- Evaluated existing and created a new framework for imputing missing data with high accuracy.- Winning project of BCSS TechDays Meshup, Hackathons, Conference Speaker.
