Advisory · Enterprise AI strategy
AI strategy for companies that want it to work
I advise leadership teams on enterprise AI strategy and agentic AI: where it creates real operational value, what to build or buy, and how to measure whether it worked.
I have spent more than a decade building AI products and technology companies, through machine learning, NLP, generative AI and now agentic systems. I co-founded CanopyLAB and the agentic AI agency Pavoreal.ai. That operator's seat is the point of view I bring: not hype cycles, but what changes the way a company actually works.
What the work covers
From diagnosis to measurable value
Diagnostics
An honest read of where AI already touches the business, where it is theatre, and which processes carry enough cost, volume or risk to be worth changing.
Prioritization
A short list of use cases ranked by business value, feasibility and the cost of being wrong, so leadership can stop funding everything a little.
Build versus buy
Clear calls on which capabilities to license, which to build, and which to wait on, grounded in how fast the underlying models are moving.
Failed AI projects
A reset for pilots that stalled: what went wrong, what is worth keeping, and whether the problem should be solved at all.
Agentic implementation
Designing agents and workflows that do real work, with the hand-offs, controls and human judgment written in from the start.
Measurable value
Agreeing the numbers before anything is built, then tracking them, so AI is judged on what it changes rather than how impressive it looks.
Questions leaders ask
- What is an enterprise AI strategy?
- A decision about where AI changes the economics of your business, in what order, and who owns the result. It is not a list of tools or a portfolio of pilots. A good strategy names the workflows that matter, the data and judgment they depend on, and the number that should move.
- What is agentic AI strategy?
- Agentic AI means systems that carry out multi-step work on their own: reading, deciding, acting in other software and handing over to people. An agentic AI strategy decides which processes can be redesigned around agents, where a human must stay in the loop, and how to measure whether the agent actually does the work.
- Why do so many AI projects fail?
- Most fail before the model is chosen. The problem was a demo rather than a process, nobody owned the outcome, the data did not reflect how work really happens, or success was never defined. The technology is rarely the first thing that breaks.
- Should we build or buy?
- Buy where the capability is becoming a commodity and the vendor will improve faster than you can. Build where the workflow, the data or the judgment is what makes you different. The distance between wanting software and making it is shrinking, which changes that answer more often than most roadmaps assume.
The thinking behind the advice
The essays are where I work out the arguments in public.
- The iPhone Killed the Calculator. Now the LLM Is Coming for Your Software Company.When the iPhone arrived, it did not have to be a better calculator, camera or map to stop people from buying those things. The LLM is about to do the same to software.
- When Your Customer Becomes Your Competitor: Your Customer Found a Shortcut to the Software FactoryPart four of how 90% of edtech disappears. Software companies spent twenty years convincing customers not to build. Artificial intelligence is changing the distance between wanting software and making it.
- Nobody Canceled the IcemanIndustries rarely disappear because a better rival wins or everyone goes bankrupt. They disappear through what I call the Five Cs of disappearance: Capital, Compression, Container, Customer and Category. Most industries meet more than one, few meet all five, and the last is the hardest to see.
- The Anatomy of a HypeHow AI extinction became a pre-IPO story, turning a real fear into evidence of power, a barrier to competition and, eventually, a valuation.
- What If Software Has a Soul?We inspect people for who they are and software for what it does. Yet software also has biases, values and a recognizable personality, and those characteristics affect the outcomes it produces.
- Elon Musk Bought the Wrong Real Estate for the Everything AppMusk bought Twitter to build the everything app. But while he was accumulating attention, OpenAI was accumulating intent, and that may be the most valuable address on the internet.