My career has revolved around spreading AI capability to where it can have the most impact.
Across government, institutions, individuals, companies and private equity, I have found different ways to turn fast-moving AI technology into practical capability people can actually use.
Where I have worked
The settings change. The question stays the same.
How do you move AI out of the lab, the keynote or the innovation team—and into the hands of people who can create meaningful value with it?
Moving from national ambition to institutional adoption.
For the past two years, I have served as AI Advisor to H.E. Dr Abdullah bin Sharaf Alghamdi, President of the Saudi Data and AI Authority. The work has shown me that serious AI progress depends as much on institutional capability, ownership and implementation discipline as it does on technology.
The frontier matters only when institutions can absorb it.
Building the systems that help a whole organization work differently.
At UNESCO's International Centre for AI Research and Ethics, I lead the development of agentic AI infrastructure: operating an organization-wide AI brain, helping teams adopt it and creating AI-native workflows around their day-to-day work.
Adoption becomes real when AI lives inside the work, not beside it.
Turning technical complexity into capability people can use.
I have delivered practical AI education at Oxford, Cambridge, Harvard, UCL and NVIDIA GTC, after starting an AI teaching community at Imperial. Later, I helped build an education company that trains thousands of people for AI careers each year.
People learn fastest by building with the tools themselves.
Finding where AI creates measurable operating value.
I have advised private companies, executive teams, PE stakeholders and portfolio businesses on AI strategy and implementation. That work has made the recurring pattern clear: value comes from changing specific workflows, not announcing a transformation.
Start with the work, find the bottleneck and prove value in practice.
The journey
A decade spent learning how AI spreads.
From teaching and startups to national AI work and company transformation, each chapter has sharpened how I think about capability, adoption and execution.
Practical AI education
Co-founded Imperial's Machine Learning Society and built a live-coding lecture format that spread to UCL, King's, LSE, Harvard, Cambridge, and Oxford.
NVIDIA
Served as NVIDIA's Ambassador to Imperial College London, connecting the student AI community with one of the core companies behind modern accelerated computing.
Y Combinator and Silicon Valley
Built inside a Y Combinator-backed startup in Silicon Valley, gaining direct exposure to venture-backed company building and founder operating pressure.
Scaled an AI training company
Co-built AI Core into a bootstrapped AI education business with roughly 30 employees, thousands of students trained each year, millions in revenue, and a negotiated exit.
Private advisory
Advised companies, PE-backed businesses, and private-equity firms on AI strategy, value creation, and practical implementation after exiting AI Core.
Saudi national AI work
Advises senior leadership at the Saudi Data and AI Authority on national AI initiatives, prototypes, and advanced agentic software.
Conversations & speaking
I share what I am learning with people doing serious work.
I speak, teach and run practical sessions on AI adoption, AI-native organizations and turning new capability into operating value.
harry@diffusing.aiNotes on practical AI
What I think matters, without the guru noise.
Occasional writing on what is changing in AI, what genuinely works, and how people and institutions can build useful capability.