The entry point · two weeks · deliberately small
AI diagnostic. Know what to build and what to leave.
A short, senior read of the AI moves that matter for your business, and the ones to leave alone. It is deliberately small, and it is the gate everything else passes through.
Why start here
Where the numbers move.
AI activity is easy to start and hard to convert. Pilots get commissioned, a vendor trial runs, and the business is left with a set of experiments and no way to rank them. The Diagnostic exists to separate the few moves that move the numbers from the many that only look busy, before any roadmap, any pilot, any spend.
We do it with an instrument. Every part of the operation is read against the same process domains, every candidate opportunity is scored the same way, and anything already running gets a verdict of its own. That is what makes the read defensible rather than a matter of opinion.
We are blunt about what to skip. Saying plainly that a use case is not worth doing is half the value, and we would rather tell you that in week two than sell you a programme that proves it in month nine.
What we look at
Framed against the Simplified Business Process Framework.
The Diagnostic is structured against the Simplified Business Process Framework (SBPF), a seven-domain operating model based on TM Forum's Business Process Framework, the industry process map for connectivity operators. Where value leaks depends on your business, so we go deep in the two or three domains where it clearly concentrates rather than thin across all of them.
Lead to Order
From prospect to activated customer and revenue recognized.
Order to Activate
Provisioning, fulfillment, and service activation.
Detect to Correct
Network assurance, fault detection, and service restoration.
Usage to Cash
Rating, billing, collections, and revenue assurance.
Contact to Resolution
Customer service, trouble tickets, and case management.
Supplier and Wholesale
Partner management, wholesale billing, and supplier settlements.
Plan to Build
Network planning, engineering, and capacity management.
For a connectivity operator the domains run from lead to order through to usage to cash. These are the kinds of questions we work through inside them.
- Where volume enters, where it queues, and where it leaves.
- Where cases fall out into manual exception handling, how large that queue is, and what it costs.
- Where the same case gets touched by multiple systems or teams before it closes.
- Where revenue leaks between systems, and where disputes, credits, and collections pile up.
- Which decisions follow rules you could write down, and which genuinely need judgment.
- What you already know is broken that no one has had the capacity to fix.
How it runs
Two weeks. Very little of your time.
Four steps, a handful of days of our time, and a few hours of yours. The read-out date is fixed on day one.
Goals first.
One conversation with the CEO, or the most senior person who can speak for the whole business: what it has to achieve this year, what is in the way, and who owns the processes behind both.
Process interviews.
Structured conversations with the people who run the highest-value parts of the operation, framed by the process domains. Two questions drive each one: what do you do, and how do you do it. We work at the level of the process, not the individual.
Score.
Every opportunity scored on value, feasibility, return and risk. Feasibility is broken out across people, process, data and systems, because a use case fails on whichever is weakest. Anything already running gets a start, stop, continue or change verdict.
Present.
A short read-out to the leadership team: the ranked shortlist, what to leave alone, and the order we would move in.
Alongside the interviews we send a one-page data request on day one: volumes, queue sizes, exception rates and the cost lines behind them. What arrives gets used. What does not is scored on the figures your people carry in their heads, with the confidence stated. Nobody runs a data project for this.
Ready to separate the AI moves that matter from the ones that only look busy?
What you walk away with
A short, ranked read.
Concrete enough to prove the thinking. Deliberately short of the how, which is the next conversation.
A ranked shortlist.
The few moves worth making, scored on value, feasibility, return and risk, with an order-of-magnitude number against each and the order we would make them in.
A verdict on what is already running.
Start, stop, continue or change, on every pilot, trial and vendor engagement already consuming money and attention.
What to leave alone.
Named plainly, with the reason, so nobody has to spend the next year finding out.
Who runs it
An operator's read. And an independent one.
Paul Salvage built and ran technology from the operator seat: CIO of a major telco group across 24 countries, Head of R&D at a telco, and Founder and COO of an open-source wireless operator built with no vendor software or licence fees. Outside telco, CTO of a Silicon Valley scale-up, CEO of a leading no-code agency, and founder of a wealth management AI startup. Building with AI in production since 2017.
The read you get is an operator's, from someone who has designed, built and run these systems and has had to live with the recommendations on the other side of the table. It is also independent. We have no product to steer you toward and no licence deal behind the advice, so when we tell you what is worth doing, that is the only thing shaping the answer.
What happens next
From shortlist to agreed work.
Once the shortlist is agreed, scope and buy decisions come from the diagnostic read. If deeper strategy work is needed, a second round of interviews maps processes and systems in full. If the first initiative is clear, we move to design and build. Either way, you leave the Diagnostic knowing exactly where AI earns its place, and where it does not.
Where would this land for your business?
Thirty minutes, no deck, no pitch. We work out which rung you are on and where the value actually is.
Book a call
Pick a time that works. Thirty minutes to talk through where AI lands in your business.
Request more information
Leave your details and we'll send you more on the Diagnostic and how it works.