Chief AI Officer · CAIO
Chief AI Officer: the person who decides what AI does in your company.
The CAIO is the executive who turns artificial intelligence into business results: deciding where it comes in, what gets scaled, what gets dropped, how it's measured and who answers for it. In large companies it's already a common role. In most SMEs nobody fills it, and that's why AI ends up coming in piecemeal.
Here's what a CAIO does, when your company needs one and how to have that function without hiring another executive.
By Diego Serrano · Growth & Innovation Advisor · Founder & Former CEOWhy now
- of organisations surveyed by IBM now have a CAIO, in just one year.
- 26% → 76%
- of CEOs say AI success depends more on people adopting it than on the technology.
- 83%
- of enterprise generative AI pilots never show up in the profit and loss account.
- 95%
- since then, the EU AI Act has required companies that use AI to make sure their people know how to (Art. 4).
- Feb 2025
What it is (and what it isn't)
A business executive, not an engineer.
The CAIO answers to the leadership team for one thing: that AI delivers real value, under control. They connect the people who know the business with the people who know the technology, manage the AI budget and account for what it returns.
They don't replace the CTO, the CIO or the head of data: they build on them. Their job is to use technology and data to change how the company works and grows. That takes enough technical judgement not to buy hype, but above all, leadership experience.
Four roles, four questions
- CTOTechnology and productWhat do we build it with?
- CIOInternal systemsDoes what we use every day work?
- CDODataIs it reliable and protected?
- CAIOAI applied to the businessWhere does AI make us win, and at what risk?
Does your company need one?
Five signs AI is already running loose in your company.
- 1
Everyone uses their own AI tool and nobody knows with what data.
- 2
You pay for AI licences without knowing what they return.
- 3
There was a promising pilot that never made it into day-to-day work.
- 4
Nobody is in charge of the EU AI Act or of training the team.
- 5
Leadership knows AI matters, but has no time to decide where to start.
If you recognise two or more, you're not missing another tool: you're missing someone to lead AI.
What a CAIO does
Five fronts. And how I work on each one.
This is what's expected of a Chief AI Officer in any company. On the right, how I work on it with my clients.
- 01
Priorities
Picks the use cases that move the business and drops the ones that only look good. A short list, ranked by impact.
How I work on it · Strategy: roadmap and metrics, with AI built in from the start.
- 02
Return
Every euro spent on licences, consultants or development has a goal and a metric. What doesn't pay back gets cut.
How I work on it · Finance: the company's numbers show which AI pays for itself.
- 03
Governance and compliance
Which tools are used, with what data and what permissions. And the EU AI Act, in order and documented.
How I work on it · Operations and systems: processes, tools and data under control.
- 04
People and adoption
Trains the team and changes how work gets done. Without adoption, the best tool stays a demo.
How I work on it · In every project: AI is implemented with the people who'll use it, not handed to them ready-made.
- 05
Bridge to technology
Translates what leadership needs into what technical people build, in-house or suppliers, and checks that what's delivered solves the problem.
How I work on it · Execution: I coordinate the specialists on your behalf, with agreed scope and fees.
A large company hires a CAIO. An SME needs the function, not the title.
A full-time CAIO is a leadership-team salary. And in a company of 10, 30 or 80 people there isn't five days a week of AI work; there is enough to decide well, implement in order and review every month.
That's an external, or fractional, CAIO: the same responsibility, with the time your company needs at each stage. More presence at the start, a monthly review once everything is working.
My work
The CAIO function, inside everything else.
I don't treat AI as a separate project. I work on the seven areas a company grows through and, in each one, AI comes in once something already works. That's why the CAIO function isn't an add-on: it's part of how I work.
- 01
Strategy and business model
Where you are, where you're going and which decisions get you there.
- 02
Product and offer
What you sell, to whom, at what price and with what margin.
- 03
Marketing and acquisition
Positioning, website, funnels and channels that bring in clients.
- 04
Sales
Sales process, CRM (Customer Relationship Management) and retention.
- 05
Operations and systems
Processes, ERP (Enterprise Resource Planning) such as Odoo and project management.
- 06
Finance and funding
Numbers, cash flow, investment rounds and grants.
- 07
Applied AI
Where it makes sense, how it's implemented and how it's measured.
AI multiplies what already works… and also what doesn’t. Order first, then AI.
See how I helpHow I work as an external CAIO
Your first months, always in this order.
The same four steps I always work in, seen from the AI side. Scope, timelines and fees are agreed before we start.
- 1
Conversation
What AI your company uses today, who uses it, with what data and what worries you. No cost, no commitment.
- 2
Analysis
I take a careful look at people, processes, tools and data: where AI can add value and where the risk lies.
- 3
Tailored proposal
The AI roadmap ordered by return, the usage rules for the team and what we won't do. With agreed scope, timelines and fees.
- 4
Execution
The chosen use cases, working day to day, with the team trained. I do it myself or coordinate the specialists, always looking after your interests.
My work ends when everything runs without me. And if one day you hire an in-house CAIO, they'll inherit a roadmap, not a mess.
Why me
What the role demands, and where I learned it.
A good CAIO combines leadership, business, technology and people. I didn't study any of the four on a master's: I've practised them for 25 years, in my own companies and in other people's.
- Business vision and leadership experience
- I founded and ran Broomx Technologies for 11 years: from zero to over €4M in revenue, operations in 24 countries and a team of up to 30 people with a leadership structure.
- Technology and AI, seen from the inside
- Broomx was an immersive technology and AI company: its own product, clinical validation and sales to hospitals. Today AI is at the centre of my own work: I use it every day to analyse, document and build.
- Defending return in front of a board
- I raised over €3M from investors such as Wayra (Telefónica) and BStartup (Banco Sabadell), and managed the company's restructuring and sale in 2025. I know how to justify an investment, and how to cut one.
- Implementing with people, not against them
- As head of systems and AI at an 18-person studio, I moved the whole team onto a single system and brought AI into quotes and tenders: from 10 to 53 projects in parallel without growing the team.
- Translating between leadership, design and technology
- I trained as a designer and ran a studio for 25 years that worked for BMW, LEGO, Nestlé and Adidas. I know how to turn a leadership goal into something a technical team can build and people want to use.
What I'm not
I'm not a model engineer and I won't sell you a platform. I've seen technology projects come to nothing up close, some of them mine, which is why I start with the problem, not the tool. When specialised development is needed, I commission and oversee it on your behalf.
Two ways to work together
With or without an AI lead on your team.
Your external CAIO
For companies where nobody holds that role. I lead your company’s AI: conversation, analysis, tailored proposal and execution, with a monthly review once it works.
Alongside your team
If you already have someone in charge of technology or AI, or you're about to hire one, I give them the method and the judgement: a clear roadmap and support until it runs on its own.
Both start the same way: with a conversation, free of charge. If I can't really add value, I'll tell you.
Frequently asked questions about the CAIO
Does an SME really need a CAIO?
It needs the function, not necessarily the title. If your company already uses AI, and it almost certainly does, someone should decide where it adds value, what risk it takes and what it returns. In an SME an external CAIO is usually enough: more time at the start and a monthly review afterwards.
How is it different from an AI consultancy?
A consultancy usually sells a project or a tool. A CAIO answers for the whole: what gets done, what doesn’t, in what order and with what return, even if that means buying nothing. And I do it looking at the whole company, not just the technology side.
Do I need my data or technology ready?
No. I start from how you work today. Often the first step isn't AI but putting processes and data in order, so AI has something to work on.
What about the EU AI Act?
It applies in stages. Since February 2025, companies that use AI must take measures so their staff have sufficient AI literacy, and since August 2026 transparency obligations apply, such as telling people when they're talking to an AI. We review what you use, with what data and what affects you, and document it. If you need specific legal advice, I'll tell you.
How much time does an external CAIO spend?
It depends on the stage: more presence during analysis and execution, then a monthly review with metrics. Scope, timelines and fees are agreed before we start.
Do you work remotely?
Yes. I work in English, Spanish and Catalan, remotely and, when it helps, in person.
Who decides what AI does in your company today?
Let's talk about how you're using it, where the risk is and where the return is. The first conversation is free of charge and with no commitment.