AI consulting for mid-sized companies and large groups
AI consulting firm: your costed AI action plan in four weeks
Aria, our conversational AI, speaks with everyone in scope, anonymously, within three days. Our senior partners cross-check these accounts against your objectives and your data. You know which AI projects to launch, which to stop, and what each one returns after all costs.
4 weeks
From scoping to findings
Over 85%
participation across our engagements
Senior partners
They lead the analysis and present the findings
Data in France
Hosted in France
“With Spentia, the teams mapped the AI opportunities, put numbers to them and raised the doubts themselves.”
— Sandrine Cormary, CEO, Omnicom Reputation Group
Leaders believe their teams are well informed about their AI strategy. Fewer than a third of employees agree.
Companies that listen to their employees and act on their feedback are seven times more likely to be AI-mature.
Survey by the BCG Henderson Institute and Columbia Business School, August 2025, 1,399 respondents in the United States.
This gap has a cost. Leadership invests where it believes AI helps. It does not see the uses that already save its teams time, sometimes with unapproved tools. A questionnaire completed by leadership will not reveal them. Only the people who do the work can describe them. Aria gives a voice to every team in scope.
Spentia measures the gap at the source: in the executive committee and with every employee.
Top-down view: a partner meets your executive committee to cover objectives, budgets, current projects and the decisions to make.
Bottom-up view: Aria invites everyone in scope to an individual, anonymous interview, within three days. Each person describes their tasks, their use of AI, what helps them and what holds them back. Over 85% participation across our engagements.
The comparison: our senior partners cross-check both views against your documents and data. They quantify the gap and set the priorities.
Leadership
objectives, budgets, trade-offs
Top-down view
a Spentia partner
Frontline
tasks, actual practices, signals
Bottom-up view · Aria
Everyone in scope
The perception gap
Quantified, dimension by dimension: what self-reported assessments miss.
We apply AI to our own work. A traditional interview programme covers a sample; Aria invites everyone in scope. Anonymity guaranteed: only aggregated results are reported.
Four weeks, from scoping to decision
Week 1, scope
Week 2, listen
Week 3, analyse
Week 4, decide
Your AI action plan includes:
a map of AI use;
a maturity profile across six dimensions;
use cases ranked by estimated value, feasibility and prerequisites;
a 12–24-month roadmap: decisions, resources, owners, financial assumptions.
With our AI adoption offer, six months of follow-up are added: monthly reviews to track actions and compare results with the starting point.
What listening reveals, and the decisions that follow
Three situations from our programmes, anonymised. Sectors and orders of magnitude have been changed; the mechanisms are accurate.
Advanced uses missing from the official inventory
The official tool inventory missed personal use and advanced practices already present in teams. Interviews revealed where time was being saved. Leadership could identify practices to extend and complete its usage register to prepare governance and AI Act work.
A pilot blocked by the decision process
A promising pilot had stalled: nobody owned the evaluation or the decision on what came next. The assessment identified this gap in the decision process. The project resumed with a leadership sponsor, indicators, a roadmap and a deadline.
Instructions that prevented action
Marketing encouraged AI while IT prohibited the use of customer data. With no rule connecting these instructions, teams abandoned projects. The assessment clarified which data and tools were allowed, and who decides when there is doubt.
“We expected a transformation plan; we approved our own. With Spentia, the teams mapped the AI opportunities, put numbers to them and raised the doubts themselves. What remained for us were the real decisions and a monthly review to stay on course.”
— Sandrine Cormary, CEO, Omnicom Reputation Group
“Within a few weeks, every task in the firm had been described and quantified by the people doing it. Discussions based on intuition stopped. We now make evidence-based decisions, AI workstream by AI workstream, with a named owner for every commitment.”
— Eric Rousseau, President, consultancy serving local authorities
How to choose
your AI consulting firm
Does the firm start with your strategy or its own solutions?
With your decisions. Scoping sets them before any choice of tool.
How does it find out what actually happens in your teams?
Aria invites everyone in scope. Over 85% participation across our engagements.
How does it verify what people tell it?
Our partners cross-check the interviews against your documents and data.
Can it recommend an approach that does not involve AI?
Yes. Automating a broken process speeds up its flaws. Simplifying or removing a task is sometimes better than automating it.
Is it independent of software vendors?
Yes. Our advice is independent of software vendors.
How does it calculate value and total cost?
We estimate value after all costs: licences, integration, oversight and the cost of errors. Your finance team reviews it with us.
Who will actually carry out the work?
Senior partners.
Which AI decision do you need to make?
Executive management
“We have committed AI budgets. I don't know what they return.”
Know which projects to continue, stop or launch, with figures to back it up.
Transformation
“Our pilots don't scale.”
Identify what is blocking them: data, decision-making or how work is organised.
Finance
“AI business cases leave out half the costs.”
We estimate each project's value after all costs and review it with you before any investment.
IT leadership
“AI tools are circulating without approval.”
An inventory of actual uses and exposed data, to build rules teams actually follow.
Do you lead a sales team? Discover our lead-to-cash action plan.
When should you engage
an AI consulting firm?
Before investing widely in AI tools
To identify useful applications and their gains before committing budgets.When pilots multiply without demonstrated value
To focus resources on useful projects, stop the others and prepare deployment.When teams already use AI without a shared framework
To map AI use, protect data and define rules teams can apply.Before deploying agents that can take action
To define permissions, oversight, controls and stop mechanisms before going live.When leadership cannot settle priorities
To compare projects by gains, risks, data and the resources required.When a transformation changes processes
Merger, reorganisation, new business model or new information system: the right time to rethink how work is shared.
Your engagement is led by a partner
Philippe Gautier
Processes, systems and transformation
He has held senior technology and transformation roles at KPMG and Michelin. He assesses what can be automated and leads realistic implementation.
Reza Ghaem-Maghami
Strategy and innovation
He led AI projects in media performance and analytics for eight years before co-founding Spentia.
Questions from executive committees
Will my teams respond?
Will my teams respond?
Over 85% participation across our engagements. Your engagement's rate is included in the findings.
Who sees the answers?
Who sees the answers?
No one in your company sees individual answers. Only aggregated results are reported. Data is hosted in France.
Is four weeks enough for a large scope?
Is four weeks enough for a large scope?
Yes. Aria runs the interviews in parallel, from 30 to 5,000 people.
You use your own AI: are you independent?
You use your own AI: are you independent?
Aria is the listening tool for the engagement. Our advice remains independent of software vendors.
How much does an engagement cost?
How much does an engagement cost?
The price depends on the scope interviewed. We set it with you at scoping, before launch.
What happens after the findings are presented?
What happens after the findings are presented?
Your action plan sets out the decisions, owners and deadlines. We can support you in implementing it if you wish.
How do an AI consulting firm, AI agency and IT services firm differ?
How do an AI consulting firm, AI agency and IT services firm differ?
The AI consulting firm defines priorities and prepares decisions. The agency designs and develops the solution. The IT services provider integrates and maintains it at scale. These roles often complement each other on the same project.
How do you identify priority AI use cases?
How do you identify priority AI use cases?
We examine the tasks teams describe, then compare possible uses by estimated value after all costs, feasibility and dependencies. You receive priorities suited to your organisation.
How do you measure AI return on investment?
How do you measure AI return on investment?
We compare expected gains with full costs: licences, models, integration, oversight, checking and errors. The starting position provides the baseline. Assumptions are validated with finance.
How should unofficial AI use be governed?
How should unofficial AI use be governed?
Anonymous interviews map shadow AI: tools used, tasks involved and data exposed. We then define appropriate rules and tools while preserving useful practices.
How do you track gains after deployment?
How do you track gains after deployment?
We first measure time spent, volumes and lead times. Each workstream then receives quantified, dated targets tracked monthly. We also track how freed time is used: advice, customer relationships, creation or other priority activities.
Does the EU AI Act apply to every business?
Does the EU AI Act apply to every business?
Every company using AI systems in the EU is affected, to very different degrees depending on the risk level of its uses and its role — provider or deployer. Prohibited practices and the obligation to support AI literacy already apply. Requirements for high-risk systems will apply from December 2027, and from August 2028 for those embedded in products. The starting point: an inventory of uses, role by role.
Do AI agents require specific deployment support?
Do AI agents require specific deployment support?
Yes. An agent can act on your tools and data. Its permissions, human oversight and stopping conditions must be defined and tested before production use.
Which AI project do you want to settle?
In 30 minutes, a partner reviews your situation and tells you whether an AI action plan would be useful. If it would not, they will tell you that too.
Philippe Gautier
Reza Ghaem-Maghami
An example AI action plan, to prepare your executive committee decisions
Map of uses, projects ranked by estimated value after all costs, roadmap: see the deliverable before talking to us.
Composite example: company, figures and situations changed.
What does an AI consulting firm do?
An AI consulting firm helps you decide where to invest and how to deploy artificial intelligence. It assesses use cases, gains, risks, data and governance requirements. It prepares the decisions that will guide solution development and integration.
Strategy and ambition
Connect each AI project to a business objective.
Maturity and readiness
Assess readiness using observed practices and results.
Use cases and value
Identify useful applications and quantify their gains and costs.
Data and technology
Check what your IT systems can actually support.
Governance and sovereignty
Set rules for use, compliance and dependencies on suppliers.
Adoption and organisation
Define AI’s tasks and your teams’ responsibilities.
Deploying a tool is not enough to deliver results
Too many projects start with a tool and a pilot before defining the need and the return on investment.
Three errors recur: choosing the tool before the need, copying use cases without examining teams' tasks, then measuring activated licences instead of results achieved.
Choosing the right projects requires understanding the gap between AI strategy and how teams use it.
A questionnaire completed by decision-makers is not enough to understand how AI is used. It can confuse available tools with those actually used and miss unofficial uses.
These uses, known as shadow AI, show where teams have moved ahead of the rules. Mapping them identifies needs, gains and exposed data, so you can set an appropriate framework.
We compare strategy with teams’ practices.
The analysis covers three dimensions.
Three dimensions to examine before deploying AI
How AI is actually used
Which tools do teams use, for which tasks and with which data? We measure frequency of use, time saved and time spent checking or correcting responses.
How work actually gets done
We examine workloads, bottlenecks, exceptions and practices that work. Use cases come from these tasks. Deployment must improve both the tools and how work is organised.
Decision-making capacity
Who decides, who checks responses and who intervenes when things go wrong? We check responsibilities, alerts and the means to correct or stop an application.
Our approach to AI consulting
rests on five principles
Connect each use to a useful decision.
Connect each use to a useful decision.
For each use, we specify the decision or task it improves and the expected result. Tool activation rates alone do not demonstrate value.
Invite everyone within scope.
Invite everyone within scope.
Everyone is invited to an individual, confidential and anonymous interview. Participation has exceeded 85% on every engagement. We compare responses across functions to understand the full scope.
Verify the facts.
Verify the facts.
We cross-check accounts against documents, usage records and available results. Each conclusion states its supporting evidence.
Choose the right response for each task.
Choose the right response for each task.
Remove, simplify, automate, enhance with AI or defer: we compare the options. Fixing a process can create more value than adding a tool.
Calculate gains after all costs.
Calculate gains after all costs.
We deduct licences, model usage, integration, human oversight and the cost of errors. The estimated value after all costs is validated with finance.
From assessment to scale
Start with your AI maturity assessment.
In four weeks, from scoping to final presentation, Spentia’s AI maturity assessment compares leadership objectives, teams’ AI use and existing projects. You receive the barriers to remove, priority use cases and a 12–24-month roadmap.