How it works
A lightweight consulting engagement, run by an AI analyst
Overture follows the same arc a good operating-model review would — interview, model, simulate, recommend — compressed into a week and grounded in evidence from every single employee, not a sample.
Day 0
Set up in ten minutes
Create your company, paste your team list (name, email, role, salary if you want sharper economics), and send invitations. Iris introduces herself as what she is: an analyst mapping how work gets done — explicitly not a performance review.
Days 1–5
The interviews
Each person has a ~20 minute conversation, on whichever channel suits them. Iris doesn't ask "what could AI automate?" — people can't answer that. She reconstructs the work itself: what arrives on your desk, what you do with it, what systems you touch, what you hand off, what breaks, what takes disproportionately long. An internal coverage model tracks what's still missing and drills until the picture is complete.
Continuous
Extraction & reconciliation
Every completed interview is converted into a structured work model: outcomes → workflows → activities, each linked to systems, decisions, approvals and handoffs, each carrying verbatim evidence. Then the reconciler goes to work across colleagues: duplicate activities merge and corroborate each other; conflicting time estimates and approval thresholds become discrepancies — and Iris quietly goes back to the people involved to clarify.
Day 6
Scoring & scenario modelling
Each activity gets an AI transformation profile — repeatability, input/output structure, judgement, human interaction, error cost, data availability — assessed qualitatively, then scored deterministically. Code, not the model, does the arithmetic: minutes × frequency × loaded cost, adjusted by adoption and success assumptions you control. The engine then builds the three futures and prices each: capacity released, tooling required, implementation cost, residual human effort.
Day 7
The deliverable
A board-level report and an interactive organisation map. Toggle Today → AI-Capable → AI-Accelerated → AI-Native and watch the company morph: workflows consolidate, agent systems appear, human effort concentrates where humans are the optimal component. Every figure is a range with explicit assumptions, and every conclusion traces back to what your own people said.
After
From insight to implementation
Each top opportunity ships with its requirements already derived from the graph: capabilities, integrations, controls, cost bands. One click sends it to to scope properly. That's the business model — the diagnostic is free, and we earn our keep building what it finds.
Ready when you are
Best fit: companies of 20–200 people with real operational workflows — finance, claims, logistics, professional services, customer operations.
Start your diagnostic