Why We Built a Lean AI Consultancy (Not a Big 4 Practice)

Before founding Prime AI Solutions, I worked in advisory at EY and held senior roles at HSBC, Shell, NatWest, Morgan Stanley, ASOS, and Unilabs. I have seen enterprise AI transformation from inside the organisations building it and from the advisory teams helping them. I understand what the Big 4 model delivers. I also understand, from direct experience, where it consistently falls short for the businesses that make up the majority of the UK and MENA economies.

This is not an anti-Big 4 piece. The large firm model exists for good reasons and serves real needs. But it serves a specific kind of client with a specific kind of problem. Mid-market businesses in AI transformation are often not that client — and they tend to find this out after the invoice arrives, not before the engagement starts.

This is why we built Prime AI Solutions the way we did. Not as a scaled version of a Big 4 practice, but as something designed from the ground up for how mid-market AI transformation actually works.

What the Big 4 AI Practice Actually Delivers

The large advisory firms are genuinely excellent at certain things. Understanding this is important, because dismissing them entirely is as wrong as assuming they are the right choice by default.

Large-scale enterprise AI transformation requires the kind of institutional resources that only large firms can deploy: hundreds of consultants for multi-market rollouts, established relationships with ERP vendors for embedded AI implementations, methodology IP developed across thousands of similar engagements, and the brand credibility that large regulated organisations sometimes require from their advisors.

For a FTSE 100 business implementing AI across thirty countries, the Big 4 is a rational choice. For a business undergoing a high-stakes transformation where the advisory firm's brand name provides board and regulator comfort independently of the outcome, the premium is justified. For regulatory and compliance engagements where the advisory firm's signoff carries its own value, there is no lean firm substitute.

The Big 4 AI practice has also invested heavily in proprietary AI platforms, accelerators, and toolkits. Some of these are genuinely valuable. They represent accumulated implementation knowledge, pre-built components, and tested methodology. They are not always relevant to a mid-market business's specific situation, but they are real assets.

Where the Large Firm Model Falls Short for Mid-Market

The same structural features that make large firms effective for enterprise clients create predictable problems for mid-market businesses. These are not flaws — they are the inherent trade-offs of the model.

The seniority bait-and-switch

You are sold the engagement by a partner with 25 years of experience. You are delivered by a team led by someone two to four years qualified. This is not unique to AI consulting, but it is particularly consequential for AI work, where the specific experience of the person in the room matters enormously. The difference between a consultant who has overseen 20 AI implementations in finance versus one who has overseen three is not marginal — it is the difference between recognising a common failure pattern in week one and discovering it in month four.

In a lean specialist firm, the senior practitioner you engage is the practitioner who works on your account. There is no bait-and-switch because there is no team to hide behind. This creates a fundamentally different accountability structure.

Overhead-driven scope expansion

Large firm billing rates must cover significant overhead: office infrastructure, marketing, internal training programmes, partner drawings, and administrative teams. This overhead is recovered through billing rates and scope. Mid-market AI projects that a lean specialist would scope as a three-month engagement at £80,000 can become a six-month engagement at £250,000 through the same overhead logic that sustains the large firm model.

More problematically, the incentive structure in large advisory firms rewards expanding scope and extending engagements. A lean specialist firm earns its repeat business by delivering the outcome and leaving the client more capable. A large firm earns its next engagement by creating the next problem. These are different incentive structures with different outcomes for clients.

Methodology over fit

Large advisory firms have spent years developing proprietary AI methodologies. These are valuable for the firms that created them. They are not always the right framework for a specific mid-market business in a specific industry at a specific stage of AI maturity. When you engage a large firm, you get their methodology applied to your problem. When you engage a lean specialist, you get a practitioner who has absorbed multiple methodologies and applies judgment about which elements fit your situation.

This is not an argument against methodology. Structured approaches to AI transformation are genuinely valuable. It is an argument against the assumption that the methodology a large firm has productised is automatically the right one for your organisation.

The Lean Firm Advantage

The lean specialist model has genuine structural advantages for mid-market AI transformation that are not available in the large firm model, regardless of the firm's expertise.

When You Should Use a Big 4 Firm

This is not advocacy for always choosing the lean option. There are situations where the large firm is the right choice, and understanding this makes the overall picture more credible.

• 1.

  Multi-country enterprise rollout. If you are implementing AI across thirty countries with multiple ERP instances and thousands of users, you need the resource depth and global presence that only large firms have. A lean specialist cannot mobilise 200 consultants. Nor should they claim to.

• 2.

  Brand credibility is part of the deliverable. For some regulated organisations, having a Big 4 firm sign off on your AI governance framework or data strategy provides its own form of assurance to regulators and boards. If the name on the engagement letter is a deliverable in itself, a lean specialist cannot replicate this.

• 3.

  Embedded ERP AI implementation. For large-scale ERP-embedded AI (Dynamics 365, SAP, Workday at enterprise scale), the large advisory firms have dedicated practices with vendor certifications and deep product knowledge that lean specialists rarely match in breadth.

• 4.

  Procurement rules require it. Some large organisations have procurement policies that mandate suppliers above a certain revenue threshold or with specific professional indemnity coverage. If your procurement rules require it, you have no choice.

Outside these situations — and for the vast majority of mid-market UK and MENA businesses we work with — none of these conditions apply. The choice is genuinely open, and the case for a lean specialist is strong.

What We Built at Prime AI Solutions

Prime AI Solutions was designed from the start around the lean model. The background I brought from EY and subsequent roles at HSBC, Shell, NatWest, Morgan Stanley, ASOS, and Unilabs gives our work the rigour, methodology, and cross-sector experience of enterprise AI consulting. But we are built to deploy that expertise differently.

Every client engagement is led by a senior practitioner who has personally overseen AI implementations at scale. We do not send graduates with six months of AI consulting experience to run client engagements. The FCCA qualification and Big 4 background is not a marketing credential — it shapes how we approach finance function AI specifically, where accounting rigour, management reporting standards, and data governance are non-negotiable.

We scope honestly. If your use case requires a three-month engagement to deliver a meaningful outcome, we say so — and we do not expand it to six. If your use case is better served by the self-paced AI for Finance Leaders course than by a consulting engagement, we will tell you that too. Our business model requires satisfied clients who return and refer, not satisfied clients who overpay for deliverables they could have achieved with less.

We remain independent on tool selection. We have no vendor alliances with Microsoft, OpenAI, Anthropic, or any specific AI platform. When we recommend Claude Projects for a finance workflow, or n8n for an automation pipeline, or Perplexity for a research team, it is because those tools fit the use case — not because we have a commercial incentive. Our clients tend to end up with tool stacks that cost them less than what a vendor-aligned firm would have recommended.

We operate across the UK and MENA markets. Our work with finance teams in London, Edinburgh, Dubai, Abu Dhabi, and Riyadh reflects the geographic diversity of the businesses we serve, not an artificial international expansion. The MENA market specifically — where many businesses are implementing AI for the first time within Vision 2030 and equivalent national programmes — benefits from advisory that understands both the international best practice and the local regulatory context.

This is not a critique of large firms. It is an explanation of why we built something different, and why we believe mid-market businesses get better AI outcomes from us than from the alternative. If you want to explore what that looks like in practice, start with an AI audit — it is a defined, bounded engagement that gives both of us a clear picture before any larger commitment is made.

Our services