01
Consulting and delivery for SMBs in regulated sectors
Consulting and building in the same team: an audit, then AI agents and automation shipped to production across real estate, legal, medical and HR. European hosting and GDPR compliance are handled from the scoping phase, not bolted on at the end.
The right case: A use case to ship in weeks, not a transformation programme
The reservation: No multi-country delivery capacity and no change-management practice: for a group-wide transformation, take one of the eight that follow.
02
Very large-scale delivery, from framing through to run
The only player able to put thousands of consultants on a single programme and then operate the result. That is a delivery capability, not a technology lead.
The right case: Multi-country transformation programmes running over several years
The reservation: Pyramid model: the team that sells is not the team that delivers. Put names in the contract.
03
AI strategy tied to economic value, with an embedded data team
McKinsey's data arm, built to connect models to P&L decisions rather than isolated projects. The deliverable is a capital-allocation decision, not a system.
The right case: Executive committees deciding where to invest before anything gets built
The reservation: Strong on the "where" and "how much", lighter on durable production deployment.
04
Build and product design backed by a consulting engagement
BCG's tech and design arm: mixed consulting, data science and product teams that actually build rather than hand over a report.
The right case: Building an internal AI product once the strategy is settled
The reservation: Cost follows BCG's cost structure, including during build phases.
05
AI compliance, audit and governance
Where Deloitte genuinely differentiates is in connecting AI deployment to regulatory requirements — audit, internal control, EU AI Act documentation.
The right case: Regulated sectors that must document their systems as much as run them
The reservation: A compliance reflex that can weigh down projects that did not need it.
06
Hybrid environments and heavy legacy estates
The best call when the real problem is not the model but the twenty years of systems it has to connect to.
The right case: Enterprises with mainframes, legacy systems and sovereignty constraints
The reservation: Natural gravity toward the IBM ecosystem — watch this if you want to stay vendor-agnostic.
07
Risk, control and back-office transformation
Positioning close to Deloitte, with a stronger anchor in finance functions and internal control.
The right case: Finance, risk and compliance functions, in groups already using them
The reservation: Weak technological differentiation against pure-play data firms.
08
Serious software engineering and continuous delivery practice
A real, published engineering culture — the Technology Radar is its shop window. The deliverable is maintainable code, not a deck.
The right case: Technical teams that want engineers, not consultants
The reservation: Poor fit if what you need is change management or strategic arbitration.
09
A network of AI experts assembled per engagement
A collective model: senior practitioners assembled per project. Faster and cheaper than a pyramid firm on well-scoped work.
The right case: Filling a specific capability gap without hiring or signing with a large firm
The reservation: Quality depends on the team assembled for you — qualify it engagement by engagement.