Employers need clearer technical roles before cutting headcount

AI’s workforce impact is also a job-definition problem

A company can reconsider its technical headcount before clearly defining the work it needs people to do. When workforce decisions start with job losses, changing responsibilities can get less attention. Andela’s analysis of technical job descriptions points to that problem: familiar titles now carry combinations of responsibilities that employers struggle to describe.

To understand those changes, Hymel, discussing Andela’s findings, asked what was happening inside individual jobs. “There’s tons of chatter that AI is deleting jobs, AI is changing jobs, but it’s been kind of hand-wavy and anecdotal,” he said. He then asked: “If these jobs are changing, what jobs are changing? And what are the underlying skill changes that are causing this heartburn in the market?”

Those questions also concern a market Andela serves. Its business connects employers with technical talent, so it benefits when companies seek help finding people for changing roles. Its analysis shows employers how hiring requirements are shifting, while Hymel’s interpretation explains how those shifts could affect job design.

Familiar technical titles now carry broader requirements

Andela’s examination of job descriptions found changing and overlapping requirements for product managers, front- and back-end engineers, developers, and related positions. A familiar title can therefore carry new expectations about what someone should know and which decisions they should make. Hiring managers need to explain how those responsibilities fit together.

Machine learning engineers illustrate this expansion. Andela found that employers were asking them for skills associated with software architects or back-end engineers, combining machine learning expertise with capabilities from neighboring roles. A hiring team needs to explain how those capabilities will be used together so candidates understand the position’s scope.

The overlap also appears in DevOps, which connects software development with running and maintaining software. In the descriptions Andela examined, DevOps roles now required cloud engineering knowledge. Employers can treat these areas as separate specialties, so combining them changes both the work assigned to a vacancy and the qualifications used to assess candidates.

Front-end engineering extends this overlap into product decisions. Front-end engineers build the parts of software that users interact with, but Hymel described employers asking them to help decide what those parts should do. “Companies are pushing for front-end engineers to have better soft skills and be able to define and develop features,” he said.

Helping define a feature adds responsibility for deciding what to build alongside implementing it. Hymel made the connection explicit: “They want front-end engineers to dip a toe into the product management role.” Stronger interpersonal skills support this broader role because feature definition requires working with people across functions. When asking for those skills, a hiring manager therefore needs to explain which product decisions the engineer will help make.

Hymel calls these overlapping requirements “skill bleeds.” Across the machine learning, DevOps, and front-end examples, employers are combining capabilities previously associated with neighboring roles. The front-end change also affects who helps define the work, so the job-design question extends beyond acquiring AI knowledge.

New role definitions give emerging work a clearer shape

As responsibilities cross established job categories, employers can recognize work they need done while struggling to describe the position responsible for it. Hymel described the difficulty directly: “What we’ve seen out of the data here is that there’s actually a need from enterprises for new areas of work that they’re having a hard time expressing or writing job posts for.”

To describe those needs, Hymel’s team used Andela’s job-description analysis to formulate emerging roles. The team examined changing requirements, identified overlaps, and built role definitions around them. These proposed definitions give employers a way to discuss combinations of work that fit awkwardly within existing categories.

The proposed titles include DevSecOps security engineer, docs-as-code engineer, and lakehouse analytics engineer. DevSecOps brings security into software development and operations; docs-as-code applies software-development methods to documentation; and a lakehouse combines flexible data storage with structured data analysis. Each title names an emerging combination of responsibilities. An employer considering one of these definitions still needs to decide how its responsibilities would fit together in a position and what that person would own.

A further proposal, FinOps reliability engineer, makes the combination concrete. FinOps concerns the financial management of technology operations; this role would combine tracking cloud infrastructure costs, budgeting token consumption, and forecasting and modeling AI-system costs. Tokens are the units of text or other data that an AI model processes or produces, so budgeting their consumption adds an AI-specific usage cost to the role.

Together, those duties would give one position responsibility for infrastructure spending, AI usage budgets, and estimates of future AI expenses. Would they form a coherent area of ownership? A leader defining the job would need to make that decision, and the proposed role makes it easier to discuss by describing the duties as connected work.

Skill obsolescence and unmet work needs can coexist

New combinations of responsibilities can emerge while existing skills lose value. Hymel said companies previously hired people with talents expected to remain useful throughout a career, while AI is making some skills obsolete. An employer can therefore need less of one capability while needing a different combination of skills.

When requirements change quickly, the gap between available skills and required work can grow. Hymel calls this talent debt: a mismatch that builds between the capabilities employers have and those their work requires. “The pace of the technology and the skills required are maturing so quickly that employers are starting to accumulate talent debt in ways they’ve never had to deal with before, and that obviously then gets expressed into job change,” he said.

For employers facing that mismatch, identifying an obsolete skill is one part of defining the workforce they need. They also need to decide which responsibilities each position should carry. The overlaps in Andela’s analysis make this a practical job-design task: leaders must determine how capabilities fit together before judging how existing roles relate to the changed work.

Because declining skill value can coexist with unmet enterprise needs, Hymel advises companies against reflexive responses such as mass layoffs. Employers may still struggle to describe the work they need in a vacancy, so deciding headcount at that stage carries an unresolved job-design problem into the next organizational or hiring decision. Defining the work gives leaders a clearer basis for choosing the capabilities they need and how to organize them.

Job seekers face the same changing requirements from the other side. Someone whose experience fits a familiar title may find postings that keep that title while asking for a broader combination of capabilities. Changing titles add another difficulty: workers must understand both how the work has shifted and which descriptions now match their experience.

These shifts can disadvantage workers whose qualifications closely match older role definitions, especially because people have different opportunities and resources to adapt. Moving into a newly defined role can be difficult for an affected worker because emerging responsibilities may demand capabilities different from those losing value. Employers can have unmet needs while job seekers remain exposed to job loss or declining demand for their skills.

Key takeaways for decision-makers

• Define work before deciding headcount: Employers evaluating AI’s workforce impact need clear descriptions of the responsibilities and capabilities their work requires. Resolving those questions gives hiring and restructuring decisions a firmer basis.
• Make broader responsibilities explicit: Technical titles increasingly combine neighboring specialties, including machine learning with back-end engineering and front-end development with product management. Hiring managers need to specify how those skills fit together and which decisions each role owns.
• Test emerging roles for coherent ownership: Andela’s proposed roles give employers a starting point for organizing new combinations of work. A proposed FinOps reliability engineer, for example, brings cloud spending, AI token budgets, and cost forecasts together; employers still need to establish whether those duties belong in one position.
• Address talent debt alongside skill obsolescence: Declining demand for existing skills can coexist with unmet needs for new capabilities. Workforce planners need to assess both gaps and account for workers’ differing opportunities to adapt.