Hey {{first name | there}}. I am currently preparing for a couple of trips this Q4, and before I head to the airport, I wanted to share this REMINDER with you. 

In today's career notes:

  • one infra job, 5 names, none of them 3 years old

  • why pedigree screening broke here, and what replaced it

  • 16,007 new listings, and why the salaries mislead

  • the window shuts on its own DEVELOPMENT

🧠PRESENTED BY EVERYTHINGDEVOPS

This video answers that with three real projects: rebuilding infrastructure nobody understands, cutting waste from a cloud bill nobody can explain, and building private AI systems that keep sensitive company data secure.

It also gets into the rest of the job: what an actual day looks like, the tools worth learning, how much cloud engineers can make, where the career can take you, and the parts nobody mentions in the recruitment ad.

Subscribe: @EverythingDevOpsHQ for deeper dives on agents, Kubernetes, MLOps, and AI infrastructure.

🧗THE EDGE: Nobody knows what this job is called yet

Check the infra job listings this month. The same job keeps showing up under different names.

AI platform engineer. AI infrastructure engineer. LLM platform engineer.

The descriptions overlap heavily. They pay in the same range. 

Almost none of them existed three years ago.

The companies aren't confused. They're trying names on. 

The work underneath is model gateways, eval infrastructure, agent runtimes, prompt versioning, and cost and latency observability across providers. 

Platform work with a different payload.

Well, we can agree that number is somewhat marketing😂

But the truth is, training pathways haven't kept pace, so companies upskill internally or fight over a small pool. 

What they want is inter-agent communication, state management, resource allocation, observability. You recognise every item on that list.

Why it matters for you

A job with five names has no credential attached.

No cert. No bootcamp track. No canonical resume shape. No "five plus years" filter, because the thing is eighteen months old and everyone knows it.

So hiring managers can't screen on pedigree. 

They fall back on whether you've visibly thought about the problem and have the experience to take up the challenge. 

That's the one screen where 4 years of experience and public work or great network beats 11 years and silence.

It doesn't stay unfair in your favour. Titles settle. A vendor ships a badge. Someone writes the definitive roadmap.

Then the listings start asking for eight years in a three-year-old field 👀😂 

You know the drill.

The people who wrote during the unsettled phase become the reference everyone else gets measured against.

Ignore the salary numbers going around. Mostly US-anchored, and a lot of the hiring is contract. This is a positioning argument, not a payday.

You already have opinions about agent runtimes and what breaks when a model touches your control plane. Those opinions will never be worth more in public than they are right now.

ACTION ITEMS FOR YOU

  1. Pick one thing you figured out running AI workloads that wasn't in any doc. 

  1. Make a post about it anywhere on the internet. 

Reply with what you picked.

🔭VISIBILITY STACK

Stay sharp. Stay visible.

Divine Odazie
CEO, EverythingDevOps

HOW DID WE DO?

Login or Subscribe to participate