IT recruiting agency EvoTalents operates at the intersection of two sides of the AI market: every day we talk to senior AI/ML engineers choosing their next role, and to CTOs and founders trying to hire those very engineers. Recently, we asked several candidates two direct questions: what matters most to you and what is the biggest red flag. In parallel, we collected market data from the hiring side. This memo brings both perspectives together, without polish and without theory.
PART 1: WHAT CANDIDATES SAY
What matters most when choosing a role
Technical challenge and product come first. Most respondents ranked technical challenge and product architecture above compensation. One engineer put it this way: he is driven by complex problems where AI forms core business logic, where there is freedom to design robust, scalable systems rather than quick-fix patches. Another emphasized that if a company hires him for a task that is not challenging, he will not accept the offer. AI moves fast, and unchallenging work means losing competitiveness in the job market.
Compensation matters, but is never the only factor. For some, the key is the balance between compensation and job stability, because many projects require significant effort and then suddenly shut down or run out of budget. Another candidate named compensation as the single most important factor but added that a good mix of everything else matters too.
Role scope, culture, flexibility. A separate theme is the balance between traditional software engineering and applied AI within a single role. Candidates want to understand what exactly they will be doing day to day. Equally important is how well the team's working style, processes, and overall culture align with their expectations.
What repels and causes offer rejections
Fake AI. Companies rebranding as "AI-first" when their entire product is just a thin wrapper over a third-party API, with zero proprietary data strategy or custom infrastructure. Teams where "AI" exists purely as marketing hype with no real technical leadership or engineering roadmap.
Management that doesn't understand the technology. If no one in the company understands AI/IT or management lacks deep technical understanding of the use case, that is a reason to leave. One candidate noted that this was exactly why he quit his last two jobs.
Unprofessional recruiting. A generic first message, little interest in the candidate's background, irregular communication, zero updates, no transparency about the role, salary range, or next steps. One respondent summed it up: the hiring process is the first real example of the company's culture. If candidates are not treated with respect during hiring, it is difficult to believe things will improve after joining.
Early stage without runway. Experienced engineers recognize that very few startups survive, and avoid early stages where projects can abruptly end without budget.
PART 2: WHAT HIRING MANAGERS AND CTOs SEE
While candidates choose where to go, companies hiring AI engineers face their own reality. And market data from 2025-2026 shows just how tough that reality is.
The global AI talent deficit in numbers
According to ManpowerGroup's 2026 Talent Shortage Survey (39,000 employers across 41 countries), AI skills have for the first time ever become the hardest to hire for globally, overtaking traditional engineering and IT capabilities. 72% of employers worldwide report difficulty hiring qualified specialists. In hard numbers: approximately 1.6 million open AI positions against 518,000 qualified candidates, meaning 3.2 open roles for every available engineer. IDC estimates the IT talent shortage will cost organizations $5.5 trillion in losses by the end of 2026 through delayed product launches, slower innovation, and burnout among the engineers who remain.
Speed of hiring: slow means losing
The average time-to-fill for senior AI roles is 66 days, 50% longer than for non-technical positions. In some markets the situation is worse: in the Bay Area, median time-to-hire for senior engineers grew from 38 days in Q3 2025 to 67 days in Q1 2026. Meanwhile, top AI candidates accept offers within 2-3 weeks. Companies that compress the process to under 25 days from first contact to final offer (maximum three interview rounds) systematically win the competition. This is not about lowering the bar but about operational readiness: a clear process, a prepared interview panel, pre-approved compensation budget.
The compensation gap: Europe vs US remote
AI engineer salaries have jumped 25-40% since 2023. Average compensation in the US has crossed $206,000 per year, with senior-level at top companies reaching $340,000-$550,000 total comp (base + equity). In the UK, senior AI engineers earn £100,000-£150,000+, while the market average sits at £63,000-£81,000. In Western Europe, senior-level comp lags the US by 35-55%, even before accounting for equity. For European startups this creates a concrete problem: they compete for the same candidate pool with American companies offering remote positions at US-level pay. The answer is not always paying more. What often works is a combination of clear equity structure, a technically ambitious project, and a transparent career trajectory. 42% of senior AI specialists already receive more than half of their total compensation through equity or token grants.
Evaluating AI skills: legacy methods are broken
Traditional technical interviews have stopped working for AI roles. 54% of developers name the mismatch between assessment tasks and actual job responsibilities as their top complaint about technical interviews. The context has shifted dramatically: according to Google, 75% of new code is now AI-generated and reviewed by engineers. A survey of 400 engineering leaders found that AI increases engineer productivity by an average of 34%, but this gain is uneven and widens the gap between strong and weak specialists. 73% of leaders believe a strong engineer is worth at least 3x their total compensation. Companies that have redesigned their interviews to assess systems thinking, architectural decisions, and the ability to work with AI tools (Google, Canva, Shopify, and Meta already allow AI use during interviews) get more accurate signal and close positions faster.
The market has shifted from research to production
The key shift of 2025-2026: companies are no longer looking for researchers who train models. They need engineers who build production AI systems: RAG pipelines, agentic architectures, inference services under load, model drift monitoring. Only 23% of AI job postings now require an advanced degree (down from 67% in 2020). Hiring managers regularly face a gap: out of an entire candidate pipeline, only a handful have real production experience, while the rest have repackaged data science or backend backgrounds as "AI/ML Engineer" on the hype wave. Telling genuine expertise from a repackaged resume has become a critical skill for those who hire.
WHERE THE GAP IS AND WHAT TO DO ABOUT IT
When you place these two perspectives side by side, clear friction points emerge, and these are exactly where hiring breaks down.
Candidates want technical challenge. Companies sell salary. Most job descriptions for AI roles lead with compensation packages and benefits lists. Candidates look first at the technical stack, the product, and the problem. If the JD does not answer "what exactly will I build and why is it hard," the strongest candidates do not even respond to the message.
Candidates value speed of process. Companies over-insure with extra rounds. Engineers explicitly name a drawn-out hiring process as a sign of company immaturity. Market data confirms it: with an average time-to-fill of 66 days for AI roles, the best candidates are gone in 2-3 weeks. Every extra round "just in case" is not additional confidence but the risk of losing your finalist.
Candidates look for technical maturity in management. Companies undervalue this factor. Engineers quit when leadership does not understand the technology. Companies rarely include management's technical literacy in their EVP (employer value proposition). Yet this is exactly what candidates assess in their first interviews.
Candidates fear fake AI. The market has moved from research to production. Engineers seek real product challenges, not a marketing wrapper. At the same time, 77% of AI roles no longer require an advanced degree but do demand production experience with RAG, agents, and MLOps. Companies that honestly communicate the current state of their product and the real scope of work have a significant advantage in attracting and retaining talent.
Hiring AI engineers? Let's talk about your process
EvoTalents specializes in placing senior and C-level technical specialists in AI/ML, cybersecurity, defence tech, and fintech. We see both sides of the market every day: we know what motivates the best candidates, and we know where companies lose them before the first interview. If you want to cut your time-to-hire and attract engineers who stay, let's talk.
