More than half of recent AI and machine learning engineering job postings from Fortune 500 companies require skills drawn from at least two different established roles, according to research released Thursday by Andela, an AI-native talent and services platform. The study examined 47,000 engineering job postings and identified over 2,000 skills feeding into 23 emerging job titles. Among 1,832 positions primarily titled for AI or ML engineers, 53% contained skills from multiple traditional roles, the report finds.

The five most common new roles bridge distinct skill sets to address specific operational needs. MLOps pipeline engineers—who construct and operate automated systems to deploy and monitor machine-learning models—draw 46% of their skills from ML engineering, 23% from DevOps, 15% from data engineering, and 8% each from AI engineering and data science, according to Andela. LLM application engineers, who develop and assess foundational models through large language model applications, blend 48% AI engineer and 34% ML engineer skills with elements of product design, software architecture, and embedded software work. FinOps reliability engineers manage cloud infrastructure for both dependability and cost, combining 36% DevOps, 27% site reliability engineering, 18% cloud engineering, and 9% each DevSecOps and cloud solutions architecture. Docs-as-code engineers apply program management and DevOps skills to technical writing, shifting from static documentation to specification-as-code. Product frontend engineers merge roughly equal parts traditional frontend development and product management with touches of full-stack engineering, UX research, and product design.

"If you're a DevOps engineer, historically, your skill bundle might have allocated 30 to 40% of pure DevOps-required skills that are rich and specific to that role, and you have a remaining bundle that is cross-role habitable," Cory Hymel, head of research at Andela, tells The New Stack. AI can now handle some of those transferable skills, he explains, making role-specific expertise more valuable than before. The report challenges claims from "the AI salespeople of the world that AI is going to push people to be more generalist," with Hymel noting the data doesn't support that narrative. Instead, emerging roles remain specialized, each targeting a particular operational or product challenge tied to AI adoption.

Companies that haven't revised their job postings face serious hiring risks, the report warns. Generic titles like "AI engineer" or "ML engineer" now attract diluted applicant pools—often the 100 fastest responders, many using AI-generated applications. Vague descriptions lead to mismatched hires, triggering employee churn, extended ramp times when new staff realize the role differs from what was advertised, and delayed roadmaps when replacements are needed. Organizations should frame job descriptions around outcomes rather than exhaustive skill checklists, Hymel argues, starting with what they want the hire to achieve. "The cost of code is going nearer to zero," he observes, so companies should prioritize soft skills, collaboration, backlog prioritization, and deployment experience over perfect Python scores. For engineers, the report recommends pursuing roles aligned with personal strengths: those who enjoy product focus and strategy should explore product frontend engineer positions, which span the full user-facing feature lifecycle from definition to launch. Technical writers concerned about AI-generated documentation should consider docs-as-code roles, which add technical program management and DevOps skills—"you're essentially writing the specs that enable spec-driven development," Hymel notes, allowing writers to contribute software directly. The shift toward specialized hybrid roles suggests that job seekers should emphasize cross-functional skills and outcome delivery over narrow technical mastery. Employers willing to clarify what success looks like—and engineers prepared to expand beyond single-role boundaries—stand to gain the most as AI reshapes engineering work.