The title peaked and fell, the skill did not
The standalone Prompt Engineer title has declined roughly thirty percent from its 2023 peak. Over the same period prompt-related skills appeared in around seventy-eight percent of AI job postings, up from under twenty percent in early 2024.
The work did not disappear. It stopped being a job on its own.
What AI-native engineering actually describes
Writing a good prompt turned out to be the easy part. The hard part is everything around it: what gets retrieved, how it is chunked, what stays in memory between turns, which tools the model can reach, and how you know when any of it degrades.
That work is now generally called context engineering, and it is a genuine engineering discipline rather than a writing one. It looks more like building a retrieval system than crafting a sentence.
The broader term gaining ground for this is AI-native engineering, meaning engineers who treat model behaviour as a system to be built, tested and monitored rather than a feature to be added. It is a more useful description of the skill than any job title that came before it.
What this means if you are staffing AI work
Hiring for prompt skill alone is buying a component of a role. What you actually need is engineers who can build and operate the system around the model, and who treat model behaviour as something to be tested rather than admired.
The disciplines this shows up in are covered under AI engineering services.
At Azendo every specialist completes AI certification through our Talent Success and Academy programme before assignment, regardless of discipline, for this reason. AI capability is no longer a separate job function to be bought in. It is a baseline for engineers.