Senior DevOps & AI Infrastructure Engineer
I build backend systems for ML/LLM products - training pipelines, microservices, and the infra that holds it all together. I care about clean architecture as much as clean data. Interested in DevOps, distributed systems, and MLOps tooling.
Microservices
ML/LLM
Backend systems
DevOps
Data Training
9+
Years of experience
Worked with different companies, agencies and trusted by international clients and organizations.
Have something in mind?
Send a message
Latest Articles
Articles driven by a fascination with the foundations of science and computation
Observability for AI Agents: Seeing What Your System Actually Did
Traditional monitoring tells you the service is up. It can't tell you the agent just gave a customer bad advice. Here's how to close that gap.
Read Article
The Real Cost of GPU Inference on Kubernetes
Where the money leaks when you serve models on K8s, and the practical fixes that plug each hole.
Read Article
Featured Works
Fresh design systems, microservices and sytem architectures
Elias Varga
Senior DevOps & AI Infrastructure Engineer
Elias is a DevOps and AI infrastructure engineer with nine years of experience building deployment platforms, Kubernetes clusters, and the pipelines that keep production calm. He specializes in the point where classic DevOps meets machine learning: GPU serving, LLM evaluation pipelines, and observability for AI systems.
I started my career as a sysadmin who got tired of fixing the same 3 a.m. problems by hand, so I began automating them. That turned into a career of building platforms that make the right way to ship software also the easiest way.
Over the years I've moved from bash scripts and cron jobs to Terraform, Kubernetes, and GitOps, and more recently into the infrastructure behind AI products: serving models efficiently, testing prompts like code, and tracing what agents actually do in production.
Career Timeline
Kestrel Cloud
2023 to present
Senior Platform & AI Infrastructure Engineer
Brightwave Systems
2020 to 2023
Senior DevOps Engineer
Lumen Freight
2017 to 2020
DevOps Engineer
I care about three things: reliability (systems that fail gracefully and recover quickly), cost (every dollar of cloud spend should have a reason), and developer experience (if engineers work around your platform, the platform is wrong). I write about what I learn, mentor junior engineers, and prefer boring, well-understood technology over clever novelty.
Outside of work, I brew coffee with far too much precision, hike with a camera, and maintain a home lab that is, at any given time, partly broken on purpose.