Skip to content
Multigrid

AI careers, skills & teams

What the work in this field actually consists of, what each role produces, and how to learn, hire, interview for and survive it without believing anybody's forecast.

Almost everything written about careers in this field is written in assertions nobody can check: which skills are in demand, what a role pays, how many of them exist, which certificate employers respect. Those claims are unverifiable at the moment of writing and wrong within a year of it, and a reader who acts on one has been badly served.

So these pages avoid the genre entirely. They describe the work rather than the title — what the person does on a Tuesday, what artefacts they produce, what a good one produces that a bad one does not. Where something is genuinely shippable they ship it: interview questions with what a strong answer contains, project briefs with acceptance criteria, a screening rubric, a triage rule for what to read. And where the answer honestly depends on your market or your situation, they say so and hand over the decision procedure instead of a recommendation.

You will not find a salary figure, a demand ranking or a growth rate anywhere in this cluster. Those numbers exist, they are local to your country and your quarter, and the pages that need them tell you where to look them up rather than inventing them.

How to Learn AI Engineering: A Path That Starts With Shipping

Five stages, each ending in a specific thing you have built and can show, with the theory pulled in at the point a stage stops working without it.

5 min read

AI Engineer vs ML Engineer vs Data Scientist: What Each Actually Does

Three roles described by the work rather than the label — what a normal Tuesday contains, what each one produces, and where the boundaries genuinely blur.

5 min read

Do You Need a Maths Background to Work in AI?

Where mathematics actually shows up in AI work, which specific concepts earn their place for each role, and a test for whether you personally need more.

4 min read

Building an AI Portfolio That Gets You Interviews

Four project briefs with acceptance criteria, chosen because each one forces a judgement call that a tutorial project never asks you to make.

6 min read

AI Engineering Interview Questions, and What a Strong Answer Contains

Nine questions constructed from the work itself, each with what it is really testing, what a strong answer includes, and the follow-up that separates understanding from recall.

5 min read

System Design Interviews for AI Systems: Three Worked Designs

Three designs worked through end to end — a grounded assistant, a bulk extraction pipeline, and an agent with permissions — with the trade-offs said out loud rather than assumed.

7 min read

The Skills That Transfer From Backend Engineering to AI

A mapping of what carries over unchanged, what is genuinely new, and the three habits from backend work that actively mislead here.

5 min read

Hiring Your First AI Engineer

How to decide what you are hiring for, what to screen for and what to ignore on a CV, plus a take-home brief and the rubric to score it with.

6 min read

Structuring an AI Team: The Functions That Must Be Owned

Eight functions every team building on models ends up needing, and how they collapse into one person, a few people, or dedicated roles as the team grows.

4 min read

Prompt Engineer: Is It a Real Job?

What the title actually named, which existing roles absorbed each part of that work, and the narrow situations where it still describes a whole job.

4 min read

Staying Current Without Drowning

A filter with an explicit test for what deserves your attention, built on categories of primary source you can verify yourself rather than a list of accounts to follow.

5 min read

Reading a Model Release Critically

How to separate the four kinds of claim in a release announcement, what is conventionally left out, and how to convert an announcement into a decision about your own system.

5 min read

Open Source Contribution as a Learning Path

Six kinds of contribution a newcomer can genuinely make to AI tooling, what each one teaches, and how to pick a first issue that will actually get merged.

5 min read

AI Certifications: Which Ones Are Worth Anything

A test you can apply to any certificate — what it proves, who verified it, and what a hiring process could substitute for it — instead of a ranking that expires.

5 min read

Freelancing and Consulting in AI

How to scope and contract work whose outcome is genuinely unknown at signing, which contract structure puts the uncertainty on which party, and how to write a scope that survives contact.

5 min read

Career Risk: Which AI Skills Will Age Badly

Sorts skills by what they are attached to — a vendor's interface, an API surface, a hardware constraint, a property of the problem — because the attachment is what decides how long a skill lasts.

5 min read

From Data Analyst to AI Engineer

What an analyst already has that most candidates do not, the five concrete gaps to close, the order to close them in, and one project that closes three at once.

4 min read

AI for Non-Engineers: What Is Actually Worth Learning

The four skills that change what a product manager, designer or operations person can do, why prompting is not one of them, and how to read a claim about a model without being able to test it.

5 min read

Mentoring Juniors in an AI-Assisted Team

The skill-acquisition problem that assistance creates, why it is a mechanism rather than a moral panic, and six practices that keep the learning while keeping the speed.

5 min read

Burnout and Pace in a Fast-Moving Field

Why the exhaustion is structural rather than personal, the arithmetic that shows keeping up was never available, and a triage rule that gives back most of the time.

5 min read

Other topics

AI careers, skills & teams · Multigrid