I’m Kerman. I work on cloud infrastructure, and the part I enjoy is never a single tool. It is the whole problem: the platform, the automation, and the person who has to live with both.
Right now most of my attention goes to where cloud engineering meets AI. Agents that can read real infrastructure and do something useful with it, MCP servers wired into the tools we already use, and the unglamorous groundwork that decides whether an agent is worth having. I wrote about that after a talk at the Plain Concepts Cloud Chapter.
Underneath there are four years of Azure, Terraform and hybrid platforms, first as an engineer and now closer to design. Almost everything I use I learned on the job or at home, on a couple of small machines that turned into the lab running this site.
Basque and Spanish are my first languages. I work in English.
The path here
Cloud engineer and cloud architect
- Azure
- hybrid and multi-cloud
- Terraform
- cloud governance
- agentic environments
- DevOps
Most of my week is design work: what the pieces should be, how they get deployed, and who owns them six months from now. I sit in planning and pre-sales conversations to shape the proposal, then stay close to delivery so the thing that gets sold is the thing that gets built.
A good part of this year went into the agentic side of that work: environments where agents can read real infrastructure, costs and policies instead of a diagram, and the internal session at the Cloud Chapter about it.
I also run technical interviews and help the engineers who join us find their feet.
what I took awayWriting the reference architecture is the easy half. The other half is three teams with a good reason to be the exception, and a design that survives them.
Master's degree in DevOps
- DevOps
- Agile
- Scrum
- multicloud
A year of distance learning while working full time. The useful part was having to write down why we did things the way we did, instead of just doing them.
Cloud engineer and site reliability engineer
- Terraform
- Packer
- Ansible
- Kubernetes
- Python
- Azure Pipelines
- on-premise and hybrid
- Scrum
An aeronautics manufacturer with a lot of machinery on the shop floor and a plan to move part of it to Azure. I was there for the whole trip, so I got to build the hybrid pieces from zero instead of inheriting them: the connection between the factory and the cloud, the automation with Terraform, Packer and Ansible, and the proof of concepts someone needed to see working before believing it.
Aerospace means many stakeholders and many partners, and most of the work was agreeing with them rather than writing the code.
what I took awayMachinery in a factory is expected to run for decades without anyone touching it, so the simple solution is the one that wins. The design that survives fifteen years without a maintenance window beats the clever one that needs a team looking after it.
Azure cloud data engineer and Power Platform architect
- Spark
- SQL
- Python
- Synapse
- Power Apps
- Power Automate
- Power BI
- GitHub Actions
My first job. I started as a cloud engineer working mostly on ETL: Spark, SQL and Python, feeding Synapse and Power BI.
Then I moved to the other end of the same stack and helped build the front people actually used, with Power Apps and the rest of Power Platform.
It was also my first contact with the cloud, and the point where I decided I wanted to be on the platform side of it instead.
what I took awayIt is the job where I decided what I wanted to be. Building the pipeline taught me what the cloud could do, and I wanted to be the one building the place it ran on.
Computer engineering, information systems
Information systems as the major. Final project on extracting relations from clinical text.
Certifications: Microsoft AZ-900 and DP-900, HashiCorp Terraform Associate, and the IBM Data Science professional certificate.
Before any of this I took part in building a digital platform for people living with mental illness, which received a grant and an award from the Basque Government in 2019.
Outside work it’s padel, the gym and cooking. The day to day version of all this lives on now.