The systems I build run across markets and languages, inside the teams that own the work, not in a central function that hands results down.
A system counts when people reach for it on an ordinary Tuesday without being reminded. Anything short of that is a pilot with better slides.
Twelve years of marketing P&L responsibility behind the AI work, so it answers to a business outcome rather than to demo quality.
What the seat
actually owns.
AI strategy & transformation
AI roadmaps and operating models, build-versus-buy decisions, and the executive and board alignment that turns a tooling budget into a transformation plan.
AI systems & automation
Custom AI agents, workflow automation, CRM and data integrations, retrieval and prompt engineering. Production systems built to outlive the pilot.
Enablement & change
Training, mentoring, prompt libraries and playbooks, and facilitation that gets tools adopted past the loudest voice in the room. Governance holds it together, below.
Not pilots.
Systems people use
every working day.
Selected work. My own ventures are named; client engagements are shared in conversation.
Lead enrichment, content and reporting, wired into the CRM.
A multi-market marketing operation running on country-specific enrichment agents wired into the CRM, so each market works from cleaner, qualified data without the manual research overhead. Content production and reporting sit on the same base. Designed to run inside the existing team rather than depend on me.
Prompt libraries and reusable AI skills for daily operations.
One-off prompting turned into a shared, repeatable team capability: prompt libraries, reusable Claude skills, and regular workshops with the teams that use them. Capability transfer is the deliverable, not a side effect.
Strong editorial that could not find its audience.
A media portal with genuinely good editorial and no distribution to match it. The AI-supported content and distribution strategy turned the archive into something people arrived for, and the growth curve changed direction within months.
GreenGina, a working AI laboratory with real money on the line.
My own DTC brand of certified organic microgreens, where AI-supported ecommerce, content, and operations workflows get tested on my own P&L before they reach client work. Fulfilment runs in partnership with a sheltered workshop employing people with disabilities.
The three questions
your board will be
asked this year.
The EU AI Act is in force and phasing in. Most companies don't need a compliance department for it. They need one accountable person who can answer these three questions in writing.
Where does AI touch our business, and at what risk tier?
A living inventory of every AI use in the company, mapped to the AI Act's risk categories. Most uses are minimal risk. The point is knowing which ones are not.
Who is allowed to use what, with which data?
A written AI policy people can actually follow: approved tools, data boundaries, human-in-the-loop checkpoints, and vendor diligence with kill criteria before anything touches production.
Can our people demonstrate AI literacy?
Article 4 expects employers to ensure a sufficient level of AI literacy in staff who work with AI. Training, prompt libraries, and playbooks are how that obligation becomes a capability instead of a checkbox.
Two ways
to hold the seat.
A standing seat, one to three days a week.
How most companies engage me. A standing Chief AI Officer seat at the leadership table, embedded in your cadence: strategy, systems, governance, and adoption, on a six-month minimum that is designed to run continuously.
See the retainer →The same mandate, as your Chief AI Officer.
For a company where AI transformation is the central bet, I take the seat full-time. Same method, full commitment: full-time in the room, building with the team, accountable for what we agreed to deliver. If you are hiring a Chief AI Officer, start with a 30-minute conversation.
Talk about the mandate →One principle holds in both forms: AI is there to free the team for the work where human relationships still win, especially in B2B and premium segments. It is not there to replace people.
Why the marketing
half matters.
The other half of my background is twelve years of senior marketing leadership across B2B and B2C: automotive, FMCG, insurance, energy, and public institutions. That commercial grounding is why my AI work targets real business outcomes instead of pilots that never ship. I work embedded, not advisory: I sit in the room, build alongside the team, and stay accountable for what we agreed to deliver.
About the Chief AI Officer mandate.
If something else is on your mind, the contact page is the fastest way to get a useful reply.
Send a message →Q01 Is Anže Frantar available for a full-time Chief AI Officer role?
Yes, for the right mandate. Anže holds the Chief AI Officer seat fractionally for companies today, and for a company where AI transformation is the central strategic bet, he takes the seat full-time.
The method is the same in both forms: set the AI strategy, build the systems, own the governance, and lead the adoption work embedded inside the team. If you are hiring a Chief AI Officer, the first step is a 30-minute conversation.
Q02 What does a Chief AI Officer actually do?
A Chief AI Officer owns four things: the AI strategy and roadmap (including build-versus-buy decisions and board alignment), the AI systems themselves (agents, automations, data integrations), AI governance (written policy, risk register, EU AI Act obligations, vendor diligence), and adoption (training, playbooks, and the change management that gets tools used past the pilot).
The common failure mode is treating the role as a technology purchase. The hard part of AI is rarely the model. It is adoption.
Q03 What evidence backs the method?
Multi-market AI systems running in production inside B2B groups: lead enrichment wired into the CRM, content production, and reporting, built so the existing teams run them without depending on the builder. Plus AI and distribution strategy work in media.
And GreenGina, Anže's own DTC brand, where every workflow is tested with real money before it reaches client work. Out of respect for the companies involved, engagements and their numbers are shared in conversation rather than published.
Q04 How does he handle the EU AI Act?
As a governance baseline, not a panic project: an inventory of AI uses mapped to the Act's risk tiers, a written AI policy with data boundaries and human-in-the-loop checkpoints, vendor diligence with kill criteria, and the staff AI literacy work that Article 4 expects from employers.
The goal is a board that can answer regulator and customer questions in writing, while teams keep shipping.