Know where AI pays off —
and where it’s liability-gated.
CAMI is an independent, expert-led AI-maturity audit for architecture, engineering and construction companies. Through one-to-one conversations with your people, CAMI establishes where AI pays off in your firm, where it must not be used, and how to capture the wins — grounded in your organisation, and mapped onto the Construction Intelligence Graph rather than a template.
From evidence to pilots
Six dimensions scored against observed evidence, and every workflow in scope passed through the liability gates, to find the pilots a firm can implement and start seeing benefits from. Stating where AI must not be used is the point of the audit.
Worked example: Contractor Ltd is a fictitious firm, and nothing shown is client data or a measured result.
Transcript
- 0:00Closer to AI-ready than you think. The audit shows how close, and what moves first.
- 0:03Six dimensions, with the evidence behind each. The engagement, a main contractor audited at portfolio level, rests on interviews across eight functions, a document-store walkthrough, a system walkthrough and a workflow inventory. The six dimensions draw as a radar; Knowledge Legibility is the binding constraint.
- 0:12Fourteen of seventeen workflows can use AI. Plotted by feasibility today against impact, they then meet the liability gates: thirteen become augment-only and three are prohibited (statutory and Building Safety Act sign-off, safety-critical method statements and permits-to-work, payment certification and entitlement decisions), each grounded in the Construction Intelligence Graph and CAMI's audit rubric. One is picked as the first pilot, because it carries no liability gate and builds Knowledge Legibility, the binding constraint: cited semantic search for commercial and claims evidence retrieval, with a target of assembling an event evidence pack in half the time, tested at eight weeks. A target, not a measured result.
- 0:25The CAMI mark. Request an audit at cami.build/contact.
Four findings, grounded in your firm and the graph
Each is established inside your firm and mapped onto the Construction Intelligence Graph, so it describes how you actually work and holds up when you act on it.
Ground truth about your firm
How your firm actually works, including the workarounds it quietly depends on, gathered from the people who do the work.
The gap between what you think you do and what you do
Evidence set against self-report. Where the two diverge is often the most useful finding.
Liability calls, reasoned and grounded
Where AI must not be used, with the statutory reasoning, grounded in the Construction Intelligence Graph and CAMI's own audit rubric. Presentable and defensible to a board.
The workflows worth automating
The workflows that pay off, named with the evidence, and the shape automation should take. The detailed build specification is separate work, and portable: your team or anyone you choose can build it.
Two weeks, inside your organisation
We talk with people across your teams and business units to understand how work really gets done, within each team and across the organisation. Below is the usual shape of an audit; we agree the details with you.
- 00Before the week
Kickoff
We agree with you the question the readout should answer, which teams and business units are in scope, and how confidentiality works.
- 01Week one
Discovery conversations
One-to-one conversations across teams and business units, to see how people work within their team and across the wider organisation. They can happen in person, on a call or in writing, whichever suits your people. Usually five to ten people, an hour each at most.
- 02Week two
Synthesis
We assess what we've heard against CAMI's audit rubric and map it onto the Construction Intelligence Graph. Out of that come the four-area shape, the heat-map, the liability calls and the first workflows worth automating. If anything is unclear, a short follow-up fills the gap.
- 03Live
Readout
We walk you through it, in person or on a call: what's distinctive about how you work, the four-area shape and the radar, the heat-map, the workflows to start with, and an honest account of where AI shouldn't be used.
- 04Some months on
Follow-up
We check in: did the first change land, and is there anything you need help with? What worked and what didn't also helps the graph learn.
One graph, one scoring frame
Every audit is mapped onto the Construction Intelligence Graph: a structured model of how construction work connects. Each one adds anonymised patterns to it, and the benchmark is read from what accumulates.
Findings are reported in four areas, with six dimensions beneath them. How they are scored is not published.
- Policy
- Technical
- Process
- People & Skills
- D1Knowledge Legibility
- D2Workflow Readiness
- D3Risk & Liability
- D4Tooling & Integration
- D5People & Adoption
- D6Commercial & Incentive
- 01
We're independent
We take no fee from any vendor we name. If a report mentions a service from our operator, Akil Intelligence Ltd, it says so.
- 02
What people tell us stays with us
Conversations are private. Your report shares what we found, never who said it, and no one is reported by name.
- 03
We keep the scoring method to ourselves
You get your results, with the evidence and the reasoning. The method stays private, so every firm is measured the same way.
- 04
Every audit helps the graph learn
Under the engagement terms, de-identified and aggregated findings join the Construction Intelligence Graph that later audits are mapped onto.
- 05
Our “must not” calls come with reasons
Every “must not” is reasoned from the law and grounded in the Construction Intelligence Graph and CAMI's audit rubric.
One method, run the same way by every assessor
CAMI and the Construction Intelligence Graph were designed as one system: the interview method, CAMI’s own audit rubric and the graph the findings are mapped onto. Assessors gather the evidence; the rubric and the graph carry the judgement. That is what lets audits run in parallel, by multiple assessors, and stay comparable over time. It is built for every role in the industry: claims, cost, programme, procurement, site management and project controls.