I'm a non-technical career changer. Is AI product management realistic?

Yes, and it is the most common successful path. AI PM needs product discovery and prioritisation at Level 3 and AI evaluation at only Level 2 — enough to judge whether model output clears the bar, not to build the system.

Are AI certifications worth it?

They certify attendance, not ability, which is why employers discount them. A verified proof of a skill an employer's spec actually lists carries more weight than a certificate, and takes 90 minutes rather than weeks.

How do I choose which role to target?

Work backwards from your existing skills rather than forwards from job titles. Look at role specs, count how many required skills you could already prove, and target the role with the smallest gap.

Guides · For talent

Changing Career Into AI Without AI Experience

You do not need an AI job title or a machine learning degree to move into AI work. Most applied AI roles require four skills — AI evaluation, prompt engineering, agent design, and LLM application architecture — and evaluation, the one employers value most, transfers directly from QA, support, analysis, and product backgrounds.

Updated 30 July 2026

Do I need a machine learning degree to work in AI?

Not for the roles that are actually hiring in volume. There are two different jobs sharing one word. Training models is a research discipline that genuinely wants deep ML background. Building products on top of models someone else trained is a different job, and it is where most of the demand is.

The second job needs system design, evaluation discipline, and product judgment. An ML degree helps surprisingly little with any of those.

Which AI skill is easiest to enter with?

AI evaluation — and it happens to be the one employers report as their biggest gap, which is an unusually good combination for a career changer.

Evaluation is defining what "correct" means for a fuzzy task, building a test set that catches real failures rather than happy paths, and being honest about what your test cannot catch. That is a QA mindset, an analyst's mindset, or a good support engineer's mindset applied to a new object.

People who move into it fastest tend to come from QA, customer support, data analysis, and operations. If you have ever built a checklist that caught the thing everyone else missed, you have the underlying instinct.

What do I do first?

Work out which skills you already have. Most career changers have two or three of the required skills already and assume they have none, because their job title does not say so.

Prove the closest one. Not the most impressive one — the one you can prove now. A single verified proof changes how you appear in a skills-based search from invisible to ranked.

Read the gap map, then close one gap. A role spec tells you exactly which skills at which levels. That converts "break into AI" from an anxiety into a two-item list.

Build one small thing and get it reviewed. A side project is unverified on a CV, but it is real evidence once the skill behind it is proven.

How long does it realistically take?

To become *visible* to skills-based employers: one challenge. Roughly 90 minutes, once.

To become *competitive* for a specific role: usually two to four proofs, because most specs have three or four required skills. That is weeks, not the year that "retrain into AI" implies — because you are proving capability you already have rather than acquiring it from zero.

To close a genuine gap where you lack the skill entirely: that is real learning time and depends on the skill. The gap map at least tells you which ones those are, instead of leaving you to study everything.

Will employers take me seriously without an AI job title?

Employers hiring on CVs mostly will not — the title filter is the whole mechanism. Employers hiring on skills routinely do, because they are looking at the same evidence for you as for everyone else.

There is also a quieter dynamic worth knowing: teams that hire on proof are actively hunting for candidates the CV screen would have deleted. A career changer with verified skills and no matching title is a candidate they can hire before their competitors can see them.

Next step

Stop rewriting your CV. Prove one skill instead.

Describe your work in plain language, take one 60–90 minute challenge, and get a shareable page of evidence. Free forever.

Start — no CV needed

Questions people ask

I'm a non-technical career changer. Is AI product management realistic?: Yes, and it is the most common successful path. AI PM needs product discovery and prioritisation at Level 3 and AI evaluation at only Level 2 — enough to judge whether model output clears the bar, not to build the system.
Are AI certifications worth it?: They certify attendance, not ability, which is why employers discount them. A verified proof of a skill an employer's spec actually lists carries more weight than a certificate, and takes 90 minutes rather than weeks.
How do I choose which role to target?: Work backwards from your existing skills rather than forwards from job titles. Look at role specs, count how many required skills you could already prove, and target the role with the smallest gap.