The market for AI consulting has exploded. Agencies, freelancers, traditional consulting firms: everyone offers services tied to artificial intelligence. The problem is that quality varies enormously, and for a company approaching AI for the first time it's hard to tell a competent consultant from someone selling smoke and mirrors.
This guide comes from our direct experience with Italian companies. We've seen AI projects go wrong because of the wrong choice of consultant, and we know what makes the difference between a successful journey and one that runs aground. Here's what to ask, what to avoid and how to evaluate.
Why you need an AI consultant (and when you don't)
An AI consultant is needed when a company wants to integrate artificial intelligence but doesn't have the in-house skills to do it effectively and safely. In practice:
- You need one when you don't know where to start and you need someone to map your processes and identify the opportunities
- You need one when you want to train the team to use AI in a structured way
- You need one when you need custom solutions that off-the-shelf tools don't cover
- You need one when you have to ensure GDPR and AI Act compliance in your use of AI
- You don't need one if the only goal is to “use ChatGPT” — that can be done with a bit of self-training and common sense
The 8 questions to ask an AI consultant
Before signing any contract, here are the questions you should ask. The answers will tell you a lot about the consultant's quality.
1. “Can you show me an AI project you've delivered for a company similar to mine?”
A serious consultant has case studies or references. They don't necessarily have to be in your exact sector, but they must be able to demonstrate experience with companies of a similar size and with comparable problems.
2. “How does your assessment process work?”
The answer should describe a structured process: process analysis, interviews with the team, identification of opportunities, prioritization. If the answer is vague or jumps straight to the “solution,” that's a red flag.
3. “What happens if AI isn't the right solution for my problem?”
An honest consultant will tell you that not every problem is solved with AI. If the answer is always “AI can do anything,” you're talking to a salesperson, not a consultant.
4. “How do you handle data security?”
The answer should touch on GDPR, the AI Act, company policies, tool selection. If the consultant doesn't bring it up spontaneously or downplays the risks, they're not the right partner to manage your data.
5. “Is team training included?”
Implementation without training is a wasted investment. The consultant must include a training plan as an integral part of the project, not as an optional “add-on.”
6. “How do you measure results?”
They must be able to define clear KPIs before the project begins: time saved, errors reduced, costs avoided, productivity increased. If they don't know how to measure success, how will you know whether it worked?
7. “Can you actually build, not just advise?”
Many consultants can talk about AI but can't build. If the need for a custom tool comes up during the assessment, the consultant must be able to deliver it or have reliable partners to do so.
8. “What happens after the consulting engagement?”
Post-implementation support is crucial. AI evolves rapidly, and the company will need updates, further training and ongoing support. Ask whether a follow-up plan is in place.
The 5 red flags: when to run
Here are the red flags that should send you looking elsewhere:
- “AI will solve all your problems”. AI is a powerful tool but it isn't the answer to everything. A consultant who promises miracles has understood neither AI nor your company.
- They immediately propose a solution without having analyzed your processes. How can they know what you need if they haven't yet understood how you work?
- They only use buzzwords and can't explain things simply. “We'll implement a machine learning framework with an end-to-end NLP pipeline” is not an answer. “We'll build a system that reads your emails and classifies them automatically” is.
- They don't talk about data security. If the consultant doesn't raise the topic of privacy and the GDPR, they're either about to land you in trouble or they don't know the subject.
- They want to lock you into long, expensive contracts from day one. The best approach is gradual: start with an assessment, implement a quick win, measure the results, then decide whether to continue.
The selection criteria: the definitive checklist
Here's a concise checklist for evaluating an AI consultant:
Has demonstrable experience with companies similar to yours
Starts from analyzing your processes, not from the technology
Is transparent about what AI can and can't do
Includes data security in their approach
Plans team training as part of the project
Knows how to measure results with defined KPIs
Can also build, not just advise
Proposes a gradual approach, not a monolithic project
Explains things in an understandable way, without needless jargon
Offers post-implementation support
AI consultant vs. generic agency: the differences
Many digital agencies have added “AI” to their services without having real expertise. Here are the differences between a specialized AI consultant and a generic agency that has tacked AI onto its price list:
| Aspect | Specialized AI consultant | Generic agency |
|---|---|---|
| Approach | Starts from business processes | Starts from the tool |
| Skills | Knows AI, development and business | Knows AI at a surface level |
| Custom development | Can build solutions | Resells third-party tools |
| Data security | Is an integral part of the service | Often overlooked |
| Outcome | Company independence | Dependence on the vendor |
In short
Choosing the right AI consultant is the first step toward a successful integration. Look for a partner who is transparent, competent and who puts your company at the center — not the technology. Be wary of anyone who promises too much and of anyone who doesn't talk about security. And above all, start with an assessment before committing to large projects: a good consultant will suggest exactly that themselves.
Want an AI consultant who meets all these criteria?
Codebaker, a consulting firm specialized in integrating artificial intelligence for Italian companies, offers free assessments, a gradual approach and expertise in both consulting and development.
Discover AI Consulting