At SAP Bulgaria, I’m helping shape an AI-powered product that identifies root causes of failed software delivery pipelines. I build asynchronous Python services and a multi-agent AI system that gathers evidence, analyzes failures, and produces actionable findings.
Before moving into my current AI Python Engineer role, I worked as a DevOps Engineer at SAP Bulgaria, designing CI/CD pipelines and automating security and observability workflows. I also built an LLM-powered troubleshooting assistant and led major incidents through root-cause analysis, recovery, and reliability improvements.
My side projects include Code Compare, an ML pipeline for measuring semantic overlap in code-review suggestions, and Codeflow, which turns requirements into validated, reviewable code changes. I also built and deployed AtelierMarie, a full-stack e-commerce platform, and am researching agentic memory for AI assistants.
