Notes And Deep Dives · Roles And Posts
Roles and posts
Job titles in IT companies are messy. This series unpacks what people actually do, where the work overlaps, and how to tell the posts apart.
Roles and posts
Hey. Let's talk about job titles.
You open a careers page and it looks like a zoo. DevOps. SRE. Platform Engineer. Solutions Engineer. AI Engineer. Same company, five names, and if you squint the descriptions all say "you will work with Kubernetes and stakeholders."
That is the problem this series exists to fix.
I am not writing HR definitions. I am writing how these posts show up in real companies. What you own. Who pages you. Where the work stops.
Why the titles fight each other
A title is a label. A post is a pile of work.
Two companies can hire a "DevOps Engineer" and mean completely different jobs. One person is writing Terraform all day. The next one is babysitting Jenkins and a ticket queue. Both are "DevOps" on LinkedIn.
The other way around happens too. A Platform Engineer and an SRE can sit in the same standup and touch the same cluster. The distinction is not the tool. The distinction is what they are paid to protect.
| If you only remember one thing | Ask this |
|---|---|
| The title | What does the careers page call it? |
| The post | What breaks if this person is on holiday? |
| The distinction | Do they build the road, drive on it, or get called when the road is on fire? |
The major splits I keep seeing
I will unpack each role in its own page. Here is the map I use before I open a job spec.
DevOps is often the glue. Pipelines, environments, the path from commit to something running. In a small company this person is also the platform, the SRE, and the one who knows why prod is sad.
SRE is reliability as a job, not a vibe. SLIs, error budgets, the 3am page, the work of making the next outage shorter than the last one.
Platform Engineer builds the internal product that other engineers consume. Clusters, golden paths, self service. If developers still open a ticket to get a namespace, the platform is not a platform yet.
SysOps / infrastructure lives closer to the metal. Hypervisors, networks, storage, the layer under the cluster. The stuff that is still there when Kubernetes is having a day off.
AI Engineer is the new noisy one. Sometimes it means models. Sometimes it means the platform that serves models. Sometimes it means a Python notebook and a hope. That mess is why this series starts there.
A startup "AI Engineer" and a bank "AI Engineer" are not the same post. Read the work, not the badge.
How to read this series
Each page is one role. I will tell you:
- What the post is supposed to own
- What it keeps stealing from the neighbour role
- What a good hire actually did last Tuesday
- What I would ask in an interview if I wanted the truth
Start with AI Engineer. That is the first one on the desk. DevOps, SRE, Platform, and the rest land here as I write them.
Do not pick a title because it sounds senior. Pick the pile of work you want to be good at. The title will keep changing. The work will not.