Every brand I talk to wants an AI strategy. What they actually mean is they want someone to tell them they're not falling behind. So they hire a consultant, run a workshop, get a 40 page roadmap with a phased implementation plan and a capability maturity model, and feel better about themselves for a quarter.

Then nothing gets built.

I know this because I've been in those rooms. 10+ years in advertising agencies across MENA, working at MullenLowe, Leo Burnett, TBWA, on campaigns for Burger King, Vimto, Hershey's, Afia, the UAE Government Media Office. I've watched the AI conversation evolve from "should we be doing something?" to "we should be doing something" to "why haven't we done something?" while the actual doing never happened.

So I started building.

What I mean by building

Not advising. Not strategising. Building.

I build AI powered marketing tools. Content automation pipelines that take a single campaign brief, generate platform specific copy, format against brand templates, and queue for review. What used to take a content team three days takes 20 minutes. The human still approves. Still edits. Still adds the thing no machine can. But the boring mechanical part is handled.

I've built trading bots that execute strategies in live markets. Agency workflow automation that connects briefing to delivery. Design engines that produce on brand visuals from text inputs. AI powered characters that live on websites and actually know what they're talking about (there's one on my portfolio site right now, go say hi).

None of these started with a strategy document. They all started with a problem I was annoyed enough to solve.

Why most AI consultants get it wrong

Most people advising brands on AI come from either a tech background or a strategy background. Rarely both. And almost never from inside an actual marketing operation.

That gap matters more than people think. Because knowing what AI can do is not the same as knowing what it should do in a specific workflow.

I've been the account director sitting in the room when the content calendar is three days late, the client approvals are stuck in email hell, and everyone's asking why the social posts don't match the campaign. I know which bottlenecks are real and which ones are just bad project management pretending to be a systems problem.

When I build an AI tool, I'm not theorising about a use case. I'm fixing a thing that annoyed me personally for years.

How I actually work

Pick one problem. Not "how should our brand use AI." That question is too big and it always leads to a roadmap that makes everyone feel productive while producing nothing. Pick the one workflow that wastes the most time and produces the least differentiated output. That's your first build.

Prototype fast. I build working prototypes in days. The goal is not perfection. The goal is something real that people can interact with, break, and give feedback on. The prototypes that survive contact with actual users are worth investing in. The rest taught you something for free.

Integrate into existing tools. Nobody wants another platform. Nobody wants another login. The best AI solutions plug into whatever a team already uses. If it requires a new tool and a training session, it's dead on arrival.

Hand it over. I build, document, train the team, and leave. If the tool can't work without me, I built it wrong.

The actual opportunity

The brands and agencies that are going to win with AI aren't the ones with the most sophisticated strategy documents. They're the ones that built something, even something small, and learned from it faster than everyone else.

Start with one workflow. One tool. One problem solved. See what works. Scale that. Ignore everything else.

If you're a brand or agency sitting on an AI roadmap that hasn't produced a single working tool, you don't need a better strategy. You need someone who builds.

That's what I do. Let's talk.