Thessaloníki DevOps Meetup

In this talk we disambiguate the definitions of DevOps, MLOps and LLMOps, the skills needed and responsibilities carried by each one, and what it takes to hop from one to another. In addition, we will dive into the technical principles encompassing the aforementioned and present a stub project — this means vibe-coded — that demonstrates concepts like Data Lineage, Compute Scheduling and Pipeline Orchestration.

Domain-driven transformation that defined the microservices movement, and later the Platform Engineering shift, is now happening in the Data Discipline. Having a federated mesh instead of a monolith warehouse allows teams to have more ownership, autonomy, and flexibility. After introducing different tools that bring Software Engineering practices to Data, I will share the open source stack that Canonical has chosen for its Data Mesh Architecture, and how the team deals with operational challenges that come with dozens of microservices.

From core principles to a pattern that bootstraps your entire platform from a single file. Diomidis and Stelios will walk through real YAML, a real repo, and finish with a live demo showing how it all works in practice.

A recap of two major industry events — KubeCon and NVIDIA GTC. Ioannis and George take the stage for a laid-back session, sharing highlights, impressions, and takeaways.

One of the most powerful aspects of Kubernetes is its extensibility. However, with flexibility comes complexity. Deploying Kubernetes add-ons extends its functionality, but for most platform engineers managing and maintaining multiple clusters across different environments, this can be challenging. Easing the burden on platform administrators allows them to be more productive and creative while enabling them to perform deployments more easily, quickly, and reliably. Attendees will learn a new, easy, and creative approach to deploying applications, whether on-premises or in the cloud. Join us to take control of your application deployment journey.

Game developers need different tooling than typical DevOps teams. They think in game systems, not infrastructure primitives, and need quick answers to questions like: What build is deployed where? Which commits finished building? How are our servers performing right now? Alex will demo Deployment Overview, an internal Blazor platform built to give game developers self-service visibility into deployments, build pipelines, test results, and server performance.

Ever shipped a distributed system only to realize you're flying blind in production? This talk challenges the notion that observability is just another tool you add at the end—it's a mindset that must be designed in from the start. Pavlos will cut through the fragmented landscape of proprietary APM agents and incompatible telemetry formats to show why OpenTelemetry has become the one instrumentation layer to rule them all. Through a (not so live) demo spanning EKS and Lambda, you'll see unified tracing, metrics, and logs across your entire stack.

Dive beyond basic Lambda calls and learn how to orchestrate complex business logic using AWS Step Functions, EventBridge, DynamoDB, and S3. Includes a live demo!

How a modern engineering team uses Terraform to manage cloud infrastructure with security, speed, and consistency. Covers IaC organization, CI/CD automation, state management, and modular design.

Best practices in platform optimization focusing on platform architecture and infrastructure, compute optimization, and code engineering optimization. How FinOps principles and observability can drive efficiency.

The journey of scaling Loopcv from 20,000 to over 1.5 million users using Docker Swarm. Challenges faced along the way—from cloud provider decisions to business trade-offs.