The tech industry is inventing roles faster than companies can name them

AI job listings are often seeking skills that don’t align with their posted job title.

That’s according to a recent study from AI-native talent and services platform Andela, which analyzed 47,101 software job listings posted by companies in the Fortune 500 between March and May. Out of the 1,832 postings for AI and ML engineers, the study found, 53% ask for technical skills that were associated with two other established roles.

“The AI engineer and the ML engineer right now are two of the most ambiguous job postings that are out there,” Cory Hymel, head of research at Andela, told IT Brew. “And [companies are] hiring for these roles, but they’re not really sure what people should be doing underneath.”

Andela also observed companies hiring for data scientists, but then asking employed professionals to ship LLM agents.

“We see companies putting out these job roles and job descriptions, and then the butts get in the seats, and they’re asked to do something drastically different,” Hymel said. “And that puts the individual at a structural disadvantage.”

Skill bill. The misalignment between titles and skills in job postings is caused by skills now having a shorter half-life, according to Hymel, who estimates it takes roughly two years for new technology and skills to emerge into the mainstream. While job titles and job descriptions have always had trouble keeping pace with changing skills, Hymel said AI and the speed at which things change is exacerbating the problem.

“Individuals are having to adapt and grow faster than we’ve ever had to do in the past, but job titles and job descriptions are lagging that trend significantly,” Hymel said.

Actual new roles. Andela identified several new tech job titles that have emerged because of AI. One is the role of the MLOps pipeline engineer, who Hymel said is tasked with scaling machine-learning models to production.

“This MLOps pipeline engineer is really one to go in and say, ‘How do we take this and how do we put some guardrails around and actually take what might be an AI pilot and scale it out to the masses?’” Hymel said.

Another new role is the LLM application engineer, who bridges the AI and ML engineer and software architect role. Other budding roles include:

• Docs-as-code engineers: A role that blends technical writing with DevOps engineering and technical program management.
• Product front-end engineers: In addition to building front-end components, professionals in this role also handle product management responsibilities such as backlogs and feature priorities.
• Lakehouse analytics engineers: This position combines data engineering, data architecture, and business intelligence development.

Hymel said companies continue to hire for these new positions without using the correct job title: “They’re hiring for this work, but they’re calling it…old roles.”

How to hire for the roles you actually want. Hymel said companies need a better understanding of what skills are required for their organization’s specific workflows. He suggested that companies take inventory of current skills within their workforces to identify gaps that need to be filled.

“The second piece to that is to say, ‘Where do you want to go as a company? What are you looking to build? What is your roadmap looking like for the next …five years?” he said.