AI Hiring Research

    What AI Companies Are Actually Hiring For in 2026

    A synthesis of the role categories AI companies are opening this year, what each one is accountable for, and why each is hard to fill. Written from Recruits Lab search data and active engagements, with every number linked to its source.

    By Darren Nelson, Founder & CEO, Recruits LabPublished Last updated

    Direct answer: what are AI companies hiring for in 2026?

    AI hiring in 2026 concentrates into five role categories, plus one fast-growing non-engineering category. They share vocabulary and very little else — separate talent pools, separate interview loops, separate compensation curves.

    1. 1Applied AI / LLM product engineershired to ship AI features customers use. Retrieval, tool calling, structured output, evaluation.
    2. 2Agentic AI engineershired to make systems that plan and act reliable enough to be given real permissions.
    3. 3Machine learning engineershired to own model quality: data, training, serving, and the metric itself.
    4. 4AI research engineershired to improve the model, largely through post-training, evaluation, and multimodal work.
    5. 5AI infrastructure engineershired to make training and inference fast and affordable at scale.
    6. 6Forward deployed engineershired to get enterprise customers into production by writing integration code on site.

    AI product management is the largest non-engineering growth area in our funnel. The single most common cause of a failed AI search is picking the wrong category above, not screening the wrong candidates.

    What AI roles are most in demand in 2026?

    The table below shows year-over-year change in Recruits Lab search volume across 2025. This is proprietary, directional data from our own practice — it describes the searches our clients ran, not the whole market.

    Year-over-year change in Recruits Lab AI search volume, 2025
    Role categoryYoY change in search volume
    Applied AI / LLM engineer+47%
    AI product manager+41%
    ML platform engineer+33%
    AI research engineer (non-frontier startups)+28%
    All venture-backed AI engineering searches+38%

    Source: Recruits Lab proprietary search data, 2025. Full context, methodology notes, and compensation detail in the 2026 AI Hiring Report.

    The six roles, and what each one is actually for

    Each section below stands on its own. If you are trying to decide which requisition to open, read the "hired to" line first and ignore the title.

    Applied AI / LLM Product Engineer

    What AI engineers are startups hiring?

    Hired to: Turn a foundation model into a feature customers pay for. These engineers usually do not train models. They are accountable for whether the AI part of the product works in front of a user.

    What they own: The AI surface of the product. Retrieval, prompting, tool calling, structured output, latency, and the evaluation harness that tells the team whether a change made the feature better or worse.

    Signals employers screen for

    • Has shipped an AI feature to production users, not a demo or notebook
    • Can describe a specific evaluation they built and what it caught
    • Talks about failure modes and fallbacks before talking about model choice
    • Has an opinion on retrieval quality grounded in something they measured

    Why it is hard to fill: The title fragmented. Companies now separate AI Product Engineer, ML Platform Engineer, LLM Engineer, and AI Research Engineer with different compensation curves, so identically titled candidates are frequently not comparable.

    Applied AI and LLM engineer searches grew 47 percent year over year in Recruits Lab's 2025 search data — the largest engineering category increase in our practice. 2026 AI Hiring Report

    Agentic AI Engineer

    What is an agentic AI engineer?

    Hired to: Take an agent from an impressive demo to something that can be given real permissions against real systems. The job is mostly reliability engineering wearing an AI title.

    What they own: Systems that plan and act rather than answer. Tool permissioning, memory, retries and rollback, guardrails, and the evaluation work that makes an agent trustworthy enough to run without a human confirming every step.

    Signals employers screen for

    • Has run an agent in production with real side effects, not read-only tasks
    • Can describe the permission model and what they refused to let the agent do
    • Has built evaluations for multi-step behavior, not single-response quality
    • Talks about cost and loop control as first-class design problems

    Why it is hard to fill: Role definition is inconsistent across companies, and the pool with genuine production agent experience is small. Many candidates have built agent prototypes; far fewer have operated one under load with real permissions.

    Practitioner observation from active Recruits Lab searches: agentic scope is most often added to an existing applied AI requisition rather than opened as its own headcount, which is a common source of mis-scoped searches.

    Machine Learning Engineer

    What is the difference between an ML engineer and an AI engineer?

    Hired to: Fix or own a measurable model-quality bottleneck. Companies hire here when the problem is the model, not the interface to it.

    What they own: Model quality. Data and feature pipelines, training decisions, serving infrastructure, and the measured performance of the model itself rather than the product surface around it.

    Signals employers screen for

    • Owns a metric, and can say what it was before and after their work
    • Comfortable with data quality problems that look unglamorous
    • Has made a build-versus-fine-tune-versus-API call and can defend it

    Why it is hard to fill: Demand shifted toward platform and serving work faster than the candidate pool re-labeled itself, so strong ML platform engineers are often hidden behind generic ML titles.

    ML platform engineer demand grew 33 percent year over year in Recruits Lab's 2025 data, driven by serving infrastructure and evaluation tooling needs. 2026 AI Hiring Report

    AI Research Engineer

    Are companies hiring AI research engineers?

    Hired to: Improve the model itself, or adapt a foundation model in ways an API cannot. At non-frontier companies this usually means post-training and evaluation rather than pretraining from scratch.

    What they own: Moving research into working systems. Pretraining and post-training, reinforcement learning from human feedback, multimodal and speech work, evaluation research, and the experimental infrastructure that makes results reproducible.

    Signals employers screen for

    • Has a reproducible result, published or internal, that someone else relied on
    • Can explain a negative result and what it ruled out
    • Distinguishes clearly between pretraining, post-training, and evaluation work

    Why it is hard to fill: This is the slowest search category in our practice and the most mandate-sensitive. Candidates frequently choose research direction and compute access over cash.

    AI research engineer searches at non-frontier startups grew 28 percent year over year in Recruits Lab's 2025 data, with talent moving from academia and large platform companies. Time-to-fill averages 12 to 20 weeks. 2026 AI Hiring Report

    AI Infrastructure Engineer

    What AI infrastructure roles are in demand?

    Hired to: Make training and inference fast enough and cheap enough to keep shipping. Companies usually open this role after scale or spend becomes the constraint, not before.

    What they own: The compute layer. Distributed training, inference serving, GPU platform operations, throughput and latency, and the cost curve underneath every AI feature the company ships.

    Signals employers screen for

    • Has reduced a real inference cost or latency number and can quantify it
    • Has operated multi-node training, including the failure recovery
    • Understands the hardware well enough to explain where the time actually goes

    Why it is hard to fill: The skill set sits between systems engineering and ML, and the strongest people are usually employed at companies where compute is the core product.

    Practitioner observation from active Recruits Lab searches: infrastructure requisitions tend to follow a cost or reliability event rather than a hiring plan, which compresses the timeline available to fill them.

    Forward Deployed Engineer

    What is a forward deployed engineer?

    Hired to: Get enterprise customers to production when the product alone will not do it. Companies open this role when deals stall at implementation rather than at the demo.

    What they own: Deployment of the product inside a customer's environment, and the integration code required to make it work there. An FDE writes software, sits with the customer, and carries responsibility for the outcome on the customer's side.

    Signals employers screen for

    • Has written and shipped code inside a customer's stack, not just configured it
    • Comfortable being the technical face of the company in a customer room
    • Can describe an integration constraint they discovered on site and solved

    Why it is hard to fill: It requires two profiles that rarely coexist: genuine engineering depth and customer-facing composure. Candidates who are strong at one are usually weak at the other, and the title means different things at different companies.

    Practitioner observation from active Recruits Lab searches: FDE searches most often fail at the scoping stage, because the company has not decided whether the hire is an engineer who talks to customers or a solutions consultant who can code.

    What skills do AI companies want in 2026?

    Across categories, the screening signal has shifted from credentials to shipped work. Two observations from our 2025–2026 searches shape how we advise clients to run loops.

    • Work product beats pedigree. In our 2025 placements, the strongest engineers were frequently non-traditional candidates whose work product was the qualifying signal rather than their employer history. (Recruits Lab practitioner observation — 2026 AI Hiring Report)
    • Titles no longer describe the work. The "applied AI engineer" title has fragmented into AI Product Engineer, ML Platform Engineer, LLM Engineer, and AI Research Engineer, each with a different compensation curve. (AI Engineer Salary Guide 2026)
    • Evaluation is the scarce skill. Across applied AI and agentic searches, the ability to design an evaluation that catches real regressions separates candidates more reliably than model familiarity. (Practitioner observation — 2026 Agentic AI Hiring Report, qualitative)

    For loop design by specialization, see How to Interview AI Engineers and the skills breakdown in the 2026 AI Engineer Hiring Report.

    What are the hardest AI roles to hire?

    Ranked by time-to-fill in Recruits Lab searches, assuming a dedicated recruiter and a founder-led process. Directional figures from our practice.

    Recruits Lab time-to-fill by AI role, 2026
    RoleTypical time-to-fillPrimary constraint
    AI Research Engineer12–20 weeksMandate quality often outweighs cash
    Staff AI / ML Engineer10–16 weeksLimited supply, frequent counter-offers
    Senior AI / ML Engineer8–12 weeksDedicated recruiter, founder-led process
    AI Product Manager8–12 weeksAI-native experience scarce
    Founding AI Engineer (seed)6–12 weeksCofounder-adjacent fit critical

    Source: Recruits Lab proprietary placement data, full table and methodology in the 2026 AI Hiring Report. Agentic AI and forward deployed engineering are not in this table because we do not publish separate time-to-fill data for them; both are constrained by pool size and role definition rather than process length.

    What AI roles do startups need first?

    A practical sequence we see work at early-stage AI companies. It is an ordering heuristic drawn from our engagements, not a benchmark.

    1. First: applied AI or founding engineer. The bottleneck at this stage is shipping a product surface, not model quality. A founding AI engineer who can ship end to end covers more ground than a specialist.
    2. Second: machine learning engineer. Once there is a measurable model-quality problem that an API call cannot solve, hire someone accountable for the metric.
    3. Third: AI infrastructure. Usually triggered by cost or reliability pressure at real usage, not by headcount planning.
    4. Fourth: forward deployed engineering. Starts when enterprise deals stall at implementation rather than at the demo.
    5. Research, when the mandate justifies it. Research hiring only works when there is a genuine research question and compute to answer it. Otherwise the hire leaves.

    A fuller sequencing framework lives in the AI Hiring Blueprint and the AI Recruiting Hub. For founding-stage specifics, see Founding Engineer Recruiters.

    Compensation orientation

    Summary only. Full bands by level, geography, bonus, and equity live in the salary guides.

    • Applied AI engineers: $190K–$480K base across mid through principal levels in the United States. AI Engineer Salary Guide 2026
    • AI research engineers: $250K–$700K base from senior through principal, with equity dominating total package at frontier labs. AI Research Engineer Salary Guide 2026
    • Direction of travel: base salaries rose 12–22 percent year over year across senior and staff AI engineering roles in 2025. 2026 AI Hiring Report

    All compensation figures above are directional and reproduced from the linked Recruits Lab guides. They are not offer advice for a specific role.

    Frequently asked questions

    What are AI companies actually hiring for in 2026?

    Five role categories account for most AI hiring in 2026: applied AI and LLM product engineers who ship AI features inside a product, machine learning engineers accountable for model quality, AI research engineers working on pretraining and post-training, AI infrastructure engineers who own training and inference compute, and forward deployed engineers who deploy AI inside customer environments. Agentic AI engineering has emerged as a distinct specialization inside the applied AI category. AI product management is the largest non-engineering growth area.

    What AI roles are most in demand in 2026?

    In Recruits Lab search volume across 2025, applied AI and LLM engineer searches grew 47 percent year over year, AI product manager hiring grew 41 percent, ML platform engineer demand grew 33 percent, and AI research engineer searches at non-frontier startups grew 28 percent. Overall venture-backed AI engineering searches grew 38 percent. These are directional figures from our own placement and search data, not an industry-wide survey.

    What are the hardest AI roles to hire in 2026?

    By time-to-fill in Recruits Lab data, AI research engineer is the hardest at 12 to 20 weeks, followed by staff AI or ML engineer at 10 to 16 weeks. Senior AI engineer, AI product manager, and founding AI engineer searches generally run 6 to 12 weeks with a dedicated recruiter. Forward deployed and agentic AI roles are hard for a different reason: the candidate pool with genuine production experience is small and the role definition is inconsistent across companies.

    What skills do AI companies want in 2026?

    The consistent screening signal is shipped production work rather than credentials. For applied AI: retrieval, tool calling, structured output, and evaluation design. For agentic systems: tool permissioning, memory, guardrails, and failure-mode analysis. For research: pretraining and post-training experience, including reinforcement learning from human feedback and multimodal work. For infrastructure: distributed training, inference latency and cost optimization, and GPU platform operations. For forward deployed roles: writing integration code inside a customer environment while running the customer relationship.

    What AI roles do startups need first?

    Early-stage AI companies typically hire an applied AI or founding engineer first, because the immediate need is shipping a working product surface rather than improving model quality. Machine learning and research hires make sense once the product has a measurable model-quality bottleneck. Infrastructure hiring usually follows real scale or cost pressure, and forward deployed engineering starts when the first enterprise customers need the product installed inside their environment.

    Are companies hiring AI research engineers outside frontier labs?

    Yes. Recruits Lab recorded 28 percent year-over-year growth in AI research engineer searches at non-frontier startups in 2025. These searches are the slowest in our practice, averaging 12 to 20 weeks, and mandate quality frequently outweighs cash in the candidate's decision.

    Methodology & sources

    This page is a synthesis. It does not introduce new data. Every quantitative claim is reproduced from a published Recruits Lab source linked inline, and each claim falls into one of three categories.

    1. Recruits Lab proprietary, directional data

    Search-volume changes, time-to-fill ranges, and compensation bands derived from our own placement and search activity, 2023–2026. Directional, drawn from our client base, and not a representative survey of the AI industry. Published in the 2026 AI Hiring Report, AI Engineer Salary Guide, and AI Research Engineer Salary Guide.

    2. Practitioner observations

    Qualitative patterns from active searches, labeled as such throughout this page. These are not measured and carry no sample size. The Agentic AI, Forward Deployed Engineer, and AI Research & Foundation Model reports are explicitly qualitative and contain no compensation, headcount, or survey data.

    3. Third-party benchmarks

    Where our compensation reporting is calibrated against external sources, those sources are named in the underlying report rather than restated here. See Research Methodology for how Recruits Lab research is produced and reviewed.

    Citation and reuse

    Journalists and researchers may cite this page with attribution to Recruits Lab and a link to this URL. When citing a figure, please carry the qualifier with it — our data is directional and drawn from our own searches. For interview requests or underlying context, use the contact page or the media page.

    Suggested citation: Nelson, D. (2026). What AI Companies Are Actually Hiring For in 2026. Recruits Lab. https://recruitslab.com/what-ai-companies-are-hiring-for-2026