Tech vs HR podcast
What do companies ask from developers
About
I build backend systems, scalable microservices, and clean software architectures. My background in physics, machine learning, and software engineering shapes how I approach design, coding, testing, and documentation. I am especially interested in the practical limits of AI-generated solutions and the systems needed to make them reliable.
Currently
Software Engineering Team Lead at Up Hellas
Available for
Speaking
Selected work
PromptCache
I built a semantic caching layer to reduce redundant LLM requests and improve cost and latency in LLM systems. The project describes reducing LLM API costs by 30–70% through semantic caching.
PulseCheck
I developed a framework-agnostic health check library for modern Python services, addressing duplicated health endpoints across microservice architectures.
Tab Memory Orchestrator
I built a Chrome extension and Rust native host that manages browser tab lifecycle using real system pressure and user context.
Evora
I developed a reliability-first event processing runtime for Python microservices.
Quantum Machine Learning Perspective
My thesis code explores a quantum machine learning perspective on gravitational-wave signal classification.
An LLM Walks into General Relativity
I examined an experiment in which an LLM generated a full presentation on General Relativity, exploring where AI-generated content succeeds in structure and narrative and where it fails in scientific reasoning. The work also considers how deterministic validation and human review can improve reliability.
Work Background