About
I make AI part of how a team works.
I lead AI enablement for Marketing & Communications at G42, the Abu Dhabi AI group. I build the products and agents, structure the knowledge they run on, and bring the people along.
01
The short version
Most organisations can now buy AI capability. Far fewer can turn it into the way work gets done. Closing that gap is my job.
I spent twenty years in communications and journalism, and I have written code alongside it for most of that time. So I cover both halves of adoption: building and tuning the systems, and getting people to trust and use them.
02
What I do
Build
I design and ship purpose-built AI products: a media-training simulator for executives, a messaging platform, a meeting co-pilot. They fit into a nine-agent system for Marketing & Communications, coordinated through a central orchestrator, which I test, tune and maintain.
Structure
Agents are only as good as what they can read. I built and govern the function's knowledge base: messaging frameworks, campaign playbooks, crisis protocols and SOPs, written so people can read them and agents can parse them.
Adopt
I run hands-on agent workshops and a multi-platform AI training series with live demos, more than ten sessions so far. I'm also the first line of support when something doesn't work, which is where most adoption is won or lost.
Govern
I evaluate and pilot new AI tools on capability, security, integration and cost, and advise leadership on what to adopt. I set the review standards for agent-drafted content and run the quality control myself.
03
How I decide what AI should do
Before I build anything, I map the workflow step by step and sort every step into one of three kinds.
Automated
An agent does it end to end.
AI-assisted
An agent drafts or analyses. A person decides.
Human-led
People do it, because the judgement or the relationship is the point.
Then I define the handoffs between them, the quality bar, and the triggers that send sensitive work to a person. I measure agent output against that bar and keep redesigning the workflow as the tools improve.
04
Why a communicator
Adoption fails for human reasons more often than technical ones. People don't use what they don't understand or don't trust.
That is the problem I worked on for two decades before AI. I built G42's internal communications function from scratch, and led change communications through a bank merger and a world expo. Rolling out AI is the same work with new tools: explain it plainly, show it working, and make the first try easy.
- Positive internal comms sentiment at G42, held since 2022
- 51% → 85%
- Consecutive Great Place to Work certifications, with HR
- 3
- Intranets built and run: G42, Expo 2020, First Gulf Bank
- 3
- Employee sessions on using AI tools in daily work
- 10+
05
The role I'm looking for
I'm looking for a senior manager or lead role in AI enablement.
The work I want to own is how a function, or a whole company, adopts AI: choosing the tools, building the agents, structuring the knowledge, training the people, and setting the standards that keep the output trustworthy.
- Roles
- AI Enablement Lead · Senior Manager, AI Enablement
- Based in
- Abu Dhabi, UAE
06
Skills and tools
AI and agents
- AI enablement programmes
- Multi-agent orchestration and agent operations
- Human–AI operating model design
- Prompt engineering and prompt libraries
- AI output QA, review standards and escalation
Knowledge and operations
- Knowledge architecture for human and AI retrieval
- SOP and playbook design
- Workflow automation
- AI tool evaluation, security and governance
Adoption
- Change management
- AI training and workshops
- Stakeholder engagement
- Rollout campaigns
- Developer and internal communities
Technical
- Python and JavaScript
- LLM APIs and integrations
- ChatGPT Enterprise and Claude
- Intranet platforms (Unily)
- Web and SEO
07
Background
- Education
- BA Journalism, University of Johannesburg
- Nationality
- South African
- Also
- Founder and editor of chatgptguide.ai · Udemy instructor in Python