# What does an AI expert actually do?

> Justas Butkus is a fractional AI officer based in Vilnius, Lithuania, founder of AINORA, MB – the company behind the Ainora and Impetora brands – and a graduate of Vilnius University.

Justas Butkus is an AI expert and fractional AI officer based in Vilnius, Lithuania, working with mid-market companies and scale-ups across the UK, EU and US at two to four days a month. He builds and operates the production AI systems he advises on, rather than only advising on them.

**"AI expert" is not one job. It covers researchers, engineers, implementation specialists, governance advisors and strategy leaders – five roles with different skills, different economics and different failure modes. Most companies looking for "an AI expert" need an implementation specialist or a strategy lead, and hire a researcher.**

Canonical: https://justasbutkus.com/ai-expert/
Last updated: 2026-08-02

---

## Which kind of AI expert do you actually need?

The term is doing far too much work. These five people all get called "AI expert" and only one of them is likely to solve your problem.

**Five roles that all get called "AI expert"**

| Type | What they are good at | Hire when |
| --- | --- | --- |
| Researcher | Advancing what models can do | You are building a model, not using one. Rare outside labs. |
| ML / AI engineer | Building and shipping systems | You know what to build and need it built well. |
| Implementation specialist | Getting AI into an existing business process | The technology exists; making it work in your operation does not. |
| Governance / compliance advisor | Risk, audit, regulatory obligations | A regulator, insurer or customer is asking questions. |
| AI strategy lead | Deciding what is worth doing at all | You have budget and no ranked plan. |

The common failure is hiring the second when you needed the fifth. A strong engineer builds exactly what you asked for, and if the request was wrong, you get a well-engineered system nobody misses when it stops.

## Why is this term this confusing in the first place?

"AI expert" is not a protected title. There is no licensing body, no exam, and no professional standard behind it, unlike "solicitor" or "chartered accountant". Anyone can use it after a few months of hands-on experience, and plenty of genuinely capable people use it accurately while plenty of confident amateurs use it identically.

That is not a reason to distrust the category. It is a reason the label itself carries almost no information, and why the questions later on this page matter more than the title on someone's profile.

## What actually happens when you hire the wrong one of the five?

**Mismatches, and what they produce**

| You needed | You hired | What you get |
| --- | --- | --- |
| A strategy lead | An ML engineer | A well-engineered system solving the wrong problem |
| An implementation specialist | A researcher | An impressive capability demonstration that never ships |
| A governance advisor | A strategy lead | A shipped roadmap nobody can defend to a regulator or insurer |
| A strategy lead | A governance advisor | Clean paperwork and no clear view of what is actually worth building |
| An implementation specialist | A governance advisor | A slow, cautious process and nothing reaching production |

Every mismatch in this table produces competent work from the person hired. The failure is not skill, it is the wrong skill for the actual gap.

## Does company size change which of the five you need first?

Yes, and smaller companies often need one person doing the job of two or three of these roles out of necessity, which changes what to look for.

- **Under fifty people:** usually cannot support five specialists. Look for a strategy lead who can also implement, and treat governance as a checklist applied by that same person rather than a separate hire.
- **Fifty to five hundred:** typically the point where strategy and implementation start to separate cleanly, and where a governance advisor becomes worth engaging directly rather than folding into someone else's scope.
- **Above five hundred, or AI core to the product:** the five roles usually justify separate people, and the coordination between them becomes its own management problem.

## Can one person actually hold more than one of these roles?

Strategy and implementation combine more often and more successfully than the market's org charts suggest. Someone who both decides what is worth building and builds it removes the handoff where a plan meets an engineer who was not in the room when it was made, which is a common point where good strategy turns into a mediocre system.

Governance combines less cleanly with either. It benefits from a degree of independence from the build decision, since the person checking whether a system is safe to ship should not be the same person under pressure to ship it. A single operator can hold governance as a discipline applied to their own work, but a larger or higher-stakes deployment usually wants a second set of eyes.

## What if you need more than one role at once?

Sequence it rather than hiring all five simultaneously. Strategy comes first, because it determines what the other four are actually needed for. Implementation follows once there is a ranked plan to execute. Governance should be designed in from the start of implementation, not retrofitted once something is already live. A dedicated researcher is very rarely the right addition for a company in this size range at all.

## What gives away the wrong type of expert in the room, before you have hired anyone?

Each mismatch has a conversational tell that shows up faster than a reference check.

- **A researcher talking to a business problem** drifts toward what is theoretically possible rather than what your process actually needs.
- **An engineer talking to a strategy problem** asks for a specification before anyone has established whether the project is worth doing.
- **A governance advisor talking to a build problem** raises every risk before discussing what the system needs to do, and never gets to a recommendation.
- **A strategy lead talking to a governance problem** treats "who is accountable when it is wrong" as a detail to sort out later rather than as a design constraint.

None of these people are being dishonest. They are answering from inside the role they actually hold, which is exactly why the mismatch is easy to miss in a single meeting and expensive to discover three months in.

## How do you tell a real AI expert from a confident one?

The field is two years into a hype cycle and the supply of confident opinion vastly exceeds the supply of operating experience. Three questions separate them quickly.

1. **What have you personally put into production, and can I look at it?** Not a prototype, not a client under NDA – something running that you can inspect. This is the single most discriminating question available.
2. **What would you tell me not to do?** Anyone who cannot name work they would decline is selling capacity, not judgement.
3. **What happens when it produces a wrong answer?** Everyone who has operated a real system has an immediate, specific answer. Everyone who has not will talk about accuracy rates.

## Why do most AI projects fail, and what does that say about who to hire?

The failure rate is documented. [RAND](https://www.rand.org/pubs/research_reports/RRA2680-1.html) reports that by some estimates more than 80% of AI projects fail, roughly twice the rate of non-AI IT projects, and [MIT NANDA](https://mlq.ai/media/quarterly_decks/v0.1_State_of_AI_in_Business_2025_Report.pdf) found 95% of enterprise generative AI pilots delivering no measurable return. Neither attributes it to the technology: RAND points at the problem being misunderstood and models optimised for the wrong metric, MIT at approach rather than model quality.

All three are selection and accountability problems, not technical ones. That is why the person who can *decide* what to build usually matters more than the person who can build it – and why the strongest position is someone who can do both, because then the decision is made by someone who knows what it costs to implement.

## What do I do?

I build and operate production AI systems, and a limited number of companies engage me to own their AI direction: what gets built, in what order, under what governance, and what gets refused.

- **Two to four days a month**, a limited number of engagements at a time.
- **A fixed-scope diagnostic first**, with a written roadmap you keep either way.
- **Milestones rather than hours**, written into the engagement.
- **You hold the keys** – accounts, infrastructure and documentation are yours throughout.
- **Independence in writing** – I own an AI company and do not procure from it in an advisory role.

## Frequently asked questions

### What does an AI expert do?

It depends which of five distinct roles they occupy: researcher, engineer, implementation specialist, governance advisor or strategy lead. Most mid-market companies searching for "an AI expert" need someone who can decide what is worth building and then get it into production, not someone who advances model capability.

### How do I know if an AI expert is credible?

Ask what they have personally put into production and can show you, what work they would decline, and what happens when a system produces a wrong answer. Operators answer the third question instantly and specifically; people who have only advised talk about accuracy rates.

### Do I need an AI expert or an AI agency?

An agency builds what you specify. An expert decides what should be specified. If you already know exactly what to build and why it pays for itself, use an agency. If that is the open question, the specification is the work.

### What does an AI expert cost?

It varies enormously by role and scope, and most published figures are rate cards from firms selling the service rather than independent data. What moves the number is days per month, regulatory exposure and whether building is included.

### Can one person cover both strategy and implementation?

More often than the market's org charts suggest, and it removes the handoff where a plan meets an engineer who was not in the room when it was made. Governance combines less cleanly with either, since it benefits from some independence from the build decision.

### Is "AI expert" a protected or regulated title?

No. There is no licensing body, exam or professional standard behind it, unlike a solicitor or a chartered accountant. Anyone can use the title, which is why the title itself carries little information and the questions on this page matter more.

### Does a small company need all five roles?

Rarely. Under about fifty people, one person usually needs to combine strategy and implementation, with governance treated as a checklist that person applies rather than a separate hire. The roles typically separate cleanly somewhere between fifty and five hundred employees.

### Which of the five should we hire first?

Strategy, in almost every case, because it determines what the other roles are actually needed for. Implementation follows once there is a ranked plan, and governance should be designed in from the start rather than retrofitted afterwards.

## Related

- [The fractional AI officer role](/fractional-ai-officer/) – What ongoing accountable AI leadership involves.
- [How to choose](/how-to-choose/) – The questions worth asking, including the awkward ones.
- [Why AI pilots stall](/ai-implementation/) – The three causes, and how to tell which is yours.

## About the author

**Justas Butkus** – a fractional AI officer based in Vilnius, Lithuania, founder of AINORA, MB – the company behind the Ainora and Impetora brands – and a graduate of Vilnius University.

## If you are trying to work out which kind you need

That is usually a short conversation, and it is worth having before you write a brief or a job description.

Contact: justas@ainora.lt · [LinkedIn](https://www.linkedin.com/in/justas-butkus/)

## Other languages

- EN: https://justasbutkus.com/ai-expert/
- LT: https://justasbutkus.com/lt/di-specialistas/