Generative AI job listings grew from roughly 55 in 2021 to nearly 10,000 by May 2025 — while general software engineering roles are still down 49%. Here is what the surge is actually hiring for and how to position yourself right now.
The generative AI job market did not just grow — it went from a niche category to a mainstream hiring priority in roughly four years. Understanding what is driving that shift, and more importantly which specific skills are getting candidates through the door, is the most practical thing a working technologist can focus on right now.
A Market That Has Sorted Sharply
The overall tech job market remains deeply depressed. General software engineering listings are down 49% from the February 2020 baseline, and the broader tech sector sits roughly 36% below pre-pandemic levels, according to analysis published in July 2025. Machine learning engineer job postings, over the same period, are up 59%. These two numbers describe the same industry at the same moment. What changed is not the labor market overall — it is what companies are willing to pay to fill.
The generative AI segment tells an even more specific story. According to research tracking AI job posting trends, listings for generative AI roles started at around 55 postings in 2021 and had climbed to just under 10,000 by the spring of 2025. That kind of trajectory in under five years is unusual in any sector, and it has not slowed.
What the Job Postings Are Actually Asking For
The composition of AI job postings reveals something that surprises many job seekers who assume ML roles are primarily about training models. One in five AI job postings — 19.7% to be precise — list NLP as a required skill, making it the single most requested AI specialization right now, according to the same dataexec.io analysis. More telling is what employers emphasize beyond any specific framework: production deployment experience now outweighs familiarity with model training in how companies evaluate candidates.
This is a meaningful shift. For years, ML engineers differentiated themselves by their ability to train and fine-tune models. The market signal in 2025 is that companies already have access to capable foundation models. What they need are engineers who can take a working prototype and put it in front of real users reliably, securely, and at scale. MLOps, API integration, evaluation frameworks, and monitoring pipelines are the practical skills bridging that gap.
| Metric | ML / AI Engineer | General Software Engineer |
|---|---|---|
| Job listing change since Feb 2020 | +59% | −49% |
| Generative AI postings (2021 → May 2025) | 55 → ~10,000 | — |
| Levels.fyi avg total comp (US) | $245,000 | — |
| Robert Half 2026 mid-band median | $170,750 | — |
| Google ML engineer median | ~$333,000 | — |
Sources: Indeed Hiring Lab via pin.com, July 2025; dataexec.io, July 2025; Robert Half 2026 Salary Guide and Levels.fyi via pin.com
Entry-Level Has the Hardest Path — But Not a Dead End
The hardest place to be in this market is at the beginning of a career. Entry-level programmer employment has contracted by 27.5%, and new-graduate hiring at major tech companies has fallen by more than half, according to recent hiring data analysis. A Stanford study reported by Ars Technica found that young workers in AI-exposed fields have seen employment fall roughly 19 percentage points below peers in roles less susceptible to AI displacement. These figures are sobering, but they describe the large-company market — not the whole market.
The companies actively filling entry-level AI roles are mid-market firms and AI-native startups, not the hyperscalers dominating the headlines. These organizations are building AI product teams and frequently cannot find candidates fast enough. They also carry less rigid credentialing requirements — a portfolio of deployed projects often matters more than a specific degree pedigree.
If you are a new graduate or a career changer, this is not the moment to compete head-on for positions at the companies everyone knows by name. It is the moment to target Series A and Series B AI companies that need to ship product and do not have the luxury of extended hiring timelines.
The Compensation Picture Is Wide — Know Your Target Tier
Salary ranges for AI and ML roles span a genuinely wide band. The Robert Half 2026 Salary Guide pegs the median for mid-band AI/ML positions at $170,750. Peer-reported figures on Levels.fyi place the US average for AI engineers at $245,000, while Google ML engineers land near $333,000 at the median. Forward-deployed engineer roles — a hybrid of ML engineering and implementation consulting — carry total compensation that ranges from around $215,000 on the lower end to well above $785,000 at senior levels, depending on employer and seniority, according to compensation research from getperspective.ai.
Knowing which tier you are targeting before entering a salary conversation is not optional. Pull current Levels.fyi data for your target employer and walk into every offer discussion with a specific comparable range in hand, not a general sense of what sounds reasonable.
How to Prepare
Retarget your job search immediately. Generic job searches will surface the most competitive postings. Filter specifically for AI-native companies at growth stages, and prioritize roles that list deployment, MLOps, or evaluation frameworks in the requirements rather than model research.
Build a deployment portfolio, not just a training portfolio. Take an open-source or API-accessible model and ship something — a working application with monitoring, error handling, and a real user interface. That artifact speaks louder than any certification in a hiring conversation.
Get specific about NLP. Given that NLP skills appear in roughly one in five AI job postings, investing time in prompt engineering, retrieval-augmented generation, and text evaluation pipelines is among the highest-return skill moves available right now.
Know your compensation tier before any conversation starts. Use Levels.fyi filtered by company size and role level. The spread between a mid-market AI role and a hyperscaler position is enormous — calibrate accordingly so you do not leave money on the table or price yourself out of realistic opportunities.
Ryzlink connects engineers with AI-era roles across the US. Browse open positions or reach out to discuss your transition.
Found this useful? Share it
Comments
No comments yet — be the first to share your thoughts.