The AI operating-model diagnostic
See the AI-native version of your company — before you build it.
Deploy Iris, our AI analyst, through your organisation. She interviews every employee, reconstructs how work actually gets done, and models three credible futures — with the economics, the tooling and the roadmap to match.
Free for companies of 20–200 people · ~7 days · your team spends 20 minutes each
The real question
Not “what can AI do for my business?” but “if we rebuilt this company around AI — what would it look like, what would it cost, and what would it return?”
Generic AI readiness assessments produce use-case lists. Overture builds a functional digital twin of your organisation — every workflow, activity, system, decision and handoff — then prices the futures against it.
How it works
Interview → Model → Simulate → Recommend
Iris interviews everyone
A relaxed 20-minute conversation per person — on the web, by email, in Slack or Teams. Not a survey: she drills into how work actually happens.
We reconstruct the work graph
Answers become activities, workflows, systems, decisions and handoffs. Colleagues corroborate each other; conflicts get clarified automatically.
Every activity is analysed
Each activity gets an AI transformation profile — repeatability, judgement, risk, data — scored deterministically, never vibes.
You explore three futures
AI-Capable, AI-Accelerated, AI-Native: an interactive map and a board-level economic model with adjustable assumptions.
The deliverable
Three credible operating models — priced, not pitched
AI-Capable
Your organisation, properly equipped
Structure unchanged. Every person gets the right AI tools, knowledge access and training. Low disruption — the fastest win.
Typically 8–15% capacity released
AI-Accelerated
Workflows redesigned around AI
AI executes the repetitive chains; people supervise, handle exceptions and own relationships. This is where the economics get serious.
Typically 25–45% of mapped work addressable
AI-Native
Designed from scratch around outcomes
If this company were founded today, with agents available from day one — what would it look like? Humans only where humans are the optimal component.
A different company, fully costed
Why it's credible
An auditable model, not a hallucinated deck
Every number traces to evidence
Click any figure and follow it down: opportunity → activities → the exact employee statements behind it.
Deterministic economics
AI judges qualitative characteristics; arithmetic does the money. Hours × frequency × loaded cost, with your assumptions on sliders.
Ranges, never fake precision
Adoption and success rates are explicit. Weak evidence is labelled, not laundered into confident numbers.
Colleagues cross-check each other
When two people describe the same process differently, Iris notices — and goes back to clarify.
From a real report
Accounts payable currently consumes 3.6 FTE
AI-accelerated residual need 0.9–1.4 FTE for exceptions, supplier relationships and controls
Capacity released £102k–£128k / year
Build cost £25k–£45k · runtime £8k–£14k / yr
Confidence 84% · corroborated by 3 colleagues · 2 discrepancies resolved
Iris meets your team where they already work
Same conversation, same analyst, any channel. Employees can pause mid-interview on the web and finish over email — the state follows them.
In seven days, you'll know what your company could become.
The diagnostic is free. If you like what the model shows, can build it — every opportunity ships with its implementation scope already written. If not, you keep the clearest picture of your own operations you've ever had.
Deploy Iris in your company