Featured insight
The companies pulling away from the pack are not the ones with the most models. They are the ones that rebuilt how decisions, capital, and talent flow around AI.
Global AI consulting · 3,400 colleagues · 24 offices · since 2019
What we do
Three ways in: sector expertise, the capabilities that build AI into the operating model, and the technology practice that keeps it running.
Nine sectors, from advanced industries to public services, each with its own economics and constraints.
Explore
Strategy, data, agents, governance, security, talent — assembled per problem rather than sold as packages.
Explore
Decision-led data products, platform engineering, and the delivery discipline that makes them durable.
ExploreLatest thinking
Our published work is written by the same people who deliver the programmes. No ghostwritten thought leadership.
Report
Pilot purgatory is rarely a modelling problem. It is a measurement, ownership, and integration problem — and all three are fixable inside a quarter.
30 September 2026 · 9 min read
Article
Agents do not slot into today's processes; they dissolve them. The design question is which parts of the process you want to keep — and why.
18 September 2026 · 12 min read
Article
You do not need a perfect data estate. You need the twenty data products that carry your decisions to be trustworthy, owned, and fast.
4 September 2026 · 8 min read
Sectors
How we work
01
We anchor every opportunity to a line in the P&L and write down the baseline before anyone builds anything.
02
Ninety-day evidence gates: production usage against the agreed baseline, not a slide deck at the end of a programme.
03
The boring middle — identity, monitoring, drift, rollback — is budgeted as a first-class workstream, not overhead.
04
A named business owner, a documented playbook, and a team that can run it without us. That is the finish line.
Client impact
+4.6 pts
first-pass yield
Trajectory models put the best historical run in front of every operator shift, early enough to intervene. Nine months from first pilot to group standard.
Advanced industries · Operations
€41m
annual avoided curtailment
Network data products and forecasting models moved constraint decisions from weekly planning to day-ahead, with a published quality score on every input.
Energy & materials · Data
-38%
claims cycle time
An agentic claims triage with bounded action space and a human escalation path, backed by an evidence trail that satisfied the supervisor on first review.
Financial services · Responsible AI
Illustrative engagements. Figures shown are representative examples used for demonstration purposes.
Capabilities
Deciding where AI changes the economics of your business — and proving it within a quarter.
The twenty data products that carry your decisions — built, owned, and measured.
From automating steps to deleting handoffs — with controls that hold.
Clarity at the top is what makes speed possible further down.
Products that assume a model is present — and interfaces that make it trustworthy.
Redesign the work first, then teach into its new shape.
AI widens the attack surface and shortens the response window. Both directions are workable.
Separating the durable bets from the noise, and building the options that matter.
Our values
Our people
Global managing partner
London · AI strategy, operating model
Leads McAi's global practice on AI operating models. Twenty years advising boards on technology-driven change; previously led enterprise data at a European universal bank.
Senior partner, financial services
Stockholm · Risk, governance, credit
Works with banks and insurers on model risk, underwriting economics, and the evidence infrastructure that makes supervisory speed defensible.
Senior partner, operations
Singapore · Supply chain, planning, manufacturing
Rebuilds planning and operations for industrial and consumer clients. Specialises in turning monthly cycles into continuous, agent-assisted replanning.
Partner, energy & retail
Madrid · Data products, demand, grid
Leads decision-led data programmes across energy and retail. Her work on flexibility markets and assortment models is used by operators in six countries.
The McAi Quarterly
Where AI is actually creating value, what failed and why, and the operating-model decisions behind both.
Careers
We hire people who prefer a measured baseline to a bold claim. Consultants, engineers, designers, and researchers — in 24 offices and remotely across the EU.
Also published
Article
21 August 2026 · 7 min read
Start a conversation
We will tell you honestly whether AI is the right tool for it, what the baseline would be, and what a ninety-day proof looks like.