David runs every Beacon leadership search personally, including AI/ML and data leadership placements.
53% of US tech job postings now require AI or ML skills, up from 29% a year earlier, and AI/ML and data science postings grew 163% year over year. This is currently the hardest, most contested technical hire in the market. Here's how to actually evaluate candidates in a space full of people who've "used AI tools" but never shipped anything real.
See the interview processTwo years ago, most engineering leaders weren't screening for AI/ML skills at all. Now it's the single hardest hire in the market. Demand for AI governance skills alone is up 150% according to the AI Workforce Consortium. Supply hasn't caught up, which means the candidates who can genuinely do this work have options, and the ones who can't have learned to sound like they can.
Plenty of candidates have built a model in a Jupyter notebook. Far fewer have owned a data pipeline other teams actually depend on.
If your data has real compliance stakes, ask directly how they've handled sensitive data before, not just whether they're familiar with the concept.
A leader in this space needs to explain data quality tradeoffs to non-technical stakeholders. Jargon-only answers are a red flag.
Strong individual output doesn't automatically mean they can build or grow a team around it.
Avoid candidates whose entire AI/ML story is a personal project or a hackathon win with no production deployment behind it. Avoid anyone who can't name the specific data source, volume, or failure mode of a system they actually shipped. Avoid candidates who lean on buzzwords instead of specifics when you ask a direct technical question.
What have they actually shipped to production, not just prototyped.
Walk through a real data pipeline or model they built: the data source, the failure mode, the fix.
If relevant, ask specifically how they've handled sensitive or compliance-bound data before.
Ask their former manager what this person actually owned versus what they were near.
On our own recent search, a Clinical Data Scientist Manager role closed at $185K total comp in just 43 days, the fastest close of any engineering search we've run, when the requirement was well-defined and worked directly. AI/ML and data leadership roles broadly are commanding a real premium as demand keeps outpacing supply.
Speed is possible in this market, but only when you know exactly what you're looking for and go get it directly. Posting a job and hoping doesn't work here.
David runs every Beacon leadership search personally, including AI/ML and data leadership placements.
Book a 15-minute call, no commitment, and we'll tell you what the market looks like for your stage and stack.
See how it works