Services · 01 · Diagnose
Find your best AI starting point.
Every role in scope, scored on evidence and priced with payback and risk.
Most AI plans start with a vendor demo and end in an argument. We start with the work: what each role does, how much of it a model can carry, and what changes when it does.
You get a per role call, a costed roadmap over the horizon you set, and a redeploy path for every person whose job changes. A board can read it in one sitting.
1. Diagnose
One role, scored and priced.
Accounts Payable Clerk
Finance, 5 people. Automate in year 3.
Net value
$1.24M
Payback
31 mo
Redeployed
3.6 people
AI adoption vs. realized automation
Net value over five years, risk adjusted. Every reduction has a redeploy path.
Turn “we should do something with AI” into a costed plan.
We inventory the tasks in each role, score them on a research backed rubric, and set the result against what the role costs today. Every move carries a payback figure and a risk rating, and each is labeled forecast until it is measured.
The answer for a role may be automate, augment, keep, or re-source. Sometimes the answer is nothing, and we say so.
Start with: AI Starting Point Diagnostic. Length and price set at scoping.
The first thing to buy
AI Starting Point Diagnostic
A focused review of one business function, using your real roles and day to day work.
We look at one function, the roles involved, and the work that takes the most time. You get a clearer picture of where AI can help, what to do first, and what to explore next.
Good for
- Business owners who want a practical starting point
- Team leaders who need to focus their AI effort
- Firms that want to look at one department or workflow first
Length and price are set at scoping. No savings figure is promised before the work is scored.
What Diagnose looks like
One method, applied role by role.
Diagnose is not a deck we hand over and leave behind. It is the base every other line builds on, and it stays current as the plan changes.
- Role inventory
- Every role in scope, its tasks, and the share of time each task takes, gathered from the people who do the work.
- Evidence scoring
- Each task scored on the rubric, with the source behind every score kept on file.
- Priced moves
- Net value, payback, and risk for each move, stated as a range until it is measured.
- Redeploy paths
- Where each affected person goes next, written before any reduction is approved.
- Board report
- A client ready document your board can read without a translator.
- NovaPath Aware, as an add on
- Observed work signals from your own systems, read at team level, set against what people said they do.
You hold at the end
- A per role call: automate, augment, keep, or re-source
- A costed roadmap over the horizon you set
- A client ready report your board can read
Pairs with, sold on its own
- AI readiness assessment. Where does each team stand today, and who is ready to move first?
- NovaPath Aware. Is the work really where people say it is?
You know your business. We know what the models can carry.
We score roles every week across firms, so we know which tasks a model handles today, which need a person in the loop, and which claims do not survive contact with real work.
That pattern recognition goes into your scores. You do not have to chase every model release to know where you stand.
Line 1 of 4