Zisis Michalaris is a Senior Software Engineer at COSMOTE Telekom, working on large-scale digital platforms and distributed systems. His work focuses on backend development, architecture and building resilient services using cloud-native technologies.
He has strong expertise in the payments domain and digital channels, along with experience in developing enterprise-grade solutions. His technical background includes microservices, event-driven systems and modern platforms powered by CI/CD practices. He is actively building applications that leverage AI to enhance system capabilities and user experience, while also leading projects from design to delivery.
Zisis is passionate about system design, automation and continuous improvement. He enjoys tackling complex problems and turning them into scalable & maintainable solutions, always aiming to bridge the gap between technology and real-world impact.
Modernizing large-scale legacy systems is one of the hardest challenges in enterprise engineering especially when dealing with highly dynamic systems where capabilities and complexity have evolved organically over time. In this talk, we present a radically different approach: treating legacy code not as technical debt, but as the source of truth and executable specification.
We walk through how we transformed a complex enterprise e-commerce platform, serving hundreds of millions of visitors per year, into a modern architecture using an AI-augmented, agent-driven methodology. At the core of this approach is a multi-stage pipeline powered by AI agents and human expertise.
This session demonstrates how combining agentic AI, structured human oversight and utilizing legacy code as a source of truth can reduce timelines by multiples while increasing confidence and quality.
Searching for speaker images...