Hiring for the AI Era

1

Foundation Models

2

AI Infrastructure & Cloud

3

AI Coding & Dev Tools

4

Creative & Media AI

5

Robotics & Physical AI

6

Legal & Enterprise AI

7

AI Research

8

Healthcare & Life Sciences

9

Customer Service AI

10

Productivity & Search

ML / Applied AI Engineer

Builds AI products end-to-end - connecting models to real data, creating AI agents, and customizing models for specific tasks

The role that turns AI research into actual products customers use. It sits between the research team and the engineering team - which makes it the hardest role to write a job description for. When this role is open and unfilled, product roadmaps stall. Every Forbes AI 50 company at the application layer is hiring for it. Demand-to-supply is estimated at 3.2:1 globally.

MLOps / AI Platform Engineer

Keeps AI systems running reliably in production - deployment pipelines, performance monitoring, and cost efficiency

Think of this role as the bridge between a data scientist building a model and customers actually using it. Without MLOps, AI projects stay in pilots forever. Demand is outpacing data scientists. Total compensation ranges from $160K-$350K+ (Levels.fyi 2025), and most candidates are passive - they won't apply, they have to be sourced.

AI Research Scientist

Advances the underlying AI models - improving reasoning, safety, and capabilities through research and experimentation

The rarest profile on this list. These candidates typically have PhDs, publish their work, and receive competing offers from the best-funded AI labs in the world. Concentrated at model builders - OpenAI, Anthropic, Mistral, xAI, Thinking Machines Lab - but increasingly required elsewhere too. Demand-to-supply is estimated at 8:1. Conference sourcing and research community relationships are the only reliable pipeline.

AI Infrastructure Engineer

Builds and manages the computing infrastructure that AI models run on - hardware, networking, and scale

Builds and manages the hardware and software that AI systems run on. Less visible than ML roles but just as hard to fill - and often the bottleneck that blocks everything else from scaling. Driven by infrastructure companies like Databricks, Crusoe, SambaNova, and Together AI. Total compensation often exceeds $200K base.

Forward-Deployed / Applied AI Architect

Embeds with enterprise customers to make AI systems work in their specific environment

Part engineer, part consultant, part account executive. This role sits with customers to make AI products actually work in their environment - legal, healthcare, finance. It requires deep technical knowledge AND industry expertise, which is why candidates are so hard to find. Look for people who have moved between technical and client-facing roles, not just one or the other.

AI Product Manager

Defines what an AI product should do, how to measure whether it works, and how users experience it

Not your typical PM hire. The job description will look similar to a standard PM role, but the screening criteria are completely different - candidates need to understand how AI models fail, how to measure whether an AI feature actually works, and how to prioritize when model outputs are probabilistic. Most PM candidates won't qualify. Screen for AI project experience early.

AI Ethics & Governance Specialist

Ensures AI systems are safe, fair, and compliant - a growing function in regulated industries

Not a volume hire - few companies have a dedicated headcount for this yet. But if your company deploys AI in a regulated industry (financial services, healthcare, legal), this function is coming. Projected to grow 35% by 2027 (BLS Occupational Outlook). What to do now: identify internally who would own this if it became a compliance requirement. That conversation is easier before the reg arrives.