Choose an AI consulting company by what it hands you at the end, not by how it talks about AI. A good firm names the workflows worth changing, ranks them, says who builds the first one, decides where your data stays and leaves you able to stop. Ask every candidate the same questions and compare the written answers side by side.
#Key takeaways
- Ask what the engagement returns as a document, not what the firm believes about AI.
- The people in the sales meeting should be the people doing the work. Ask for names.
- A recommendation without a build path is a report. Ask who takes the first workflow to production.
- Data residency and approval rules belong in the scope, not in a later phase.
- Compare firms on the same written questions so the answers are comparable.
#What should an AI consulting company hand you at the end?
A ranked list of workflows, each with the reason it made the list, the cost it removes and how it would be measured. One recommended first build, with its scope and its hosting region. A governance note that says who approves what the AI produces. If a firm cannot describe its deliverable in those terms before you sign, the deliverable will be a slide deck.
This matters because most pilots never pay back. Coverage of the MIT NANDA study reported that 95% of organizations saw no business return from generative AI (virtualization review, august 2025). Gartner expected 30% of generative AI projects to be abandoned after proof of concept by the end of 2025 (Gartner, july 2024). A consulting engagement should be designed to avoid becoming one of those numbers.
#Who will actually do the work?
Ask for the names and roles of the people who will interview your team, map the process and write the recommendation. Then ask who builds the first workflow if you go ahead. A firm where senior people sell and junior people deliver is common. It is worth knowing before you start, not after.
At Stennir the people on the first call are the people who run the audit and build the first workflow. You can see who they are on the about page.
#How does the firm decide which workflow comes first?
A good answer names criteria you can check: volume, how rule-based the work is, how costly a mistake is, who would approve the output and whether the data is reachable. A weak answer starts with a model or a tool. We wrote up the criteria Stennir uses in how to pick the first process to automate with AI.
#What questions should you ask before you sign?
- What exactly do we receive at the end, and can we see an example with the client removed?
- Who on your side runs the interviews, and who builds the first workflow?
- Where will our data be processed and stored, and who decides that?
- Who on our side approves what the AI produces, and how is it logged?
- What happens to the accounts, the data and the off switch if we stop working with you?
- Which risk framework do you work from? The NIST AI risk management framework, january 2023 and ISO/iec 42001, december 2023 are reasonable answers.
We are
Stennir runs an audit that ends in ranked workflows, one recommended build with its hosting region, and the named people who would build it.
We aren't
Stennir does not sell a strategy deck that ends at the recommendation, or a staffing model where the people who sell are not the people who work.
#What should an operations lead do next?
Send the questions above to every firm on your list and compare the written answers. If you want a starting point before talking to anyone, take the free readiness assessment. If you want to see how Stennir answers each one, read AI consulting, see AI for operations, or book a 30-minute call.
#Common questions
How many AI consulting companies should we talk to before choosing?
Enough to compare answers to the same questions, which for most mid-sized companies means a short list of three or four. More than that and the comparison turns into a reading exercise. Fewer and you have no baseline for what a strong answer looks like. Send all of them the same written questions.
Should we pick a big firm or a small specialist firm?
Pick the one whose named people will do the work and whose deliverable ends in a build path. Large firms bring breadth and frameworks. Small firms usually put senior people on the work directly. The right choice depends on whether you need a programme across many units or one workflow running well first.
What do you need from us to start an AI consulting engagement?
A sponsor who can approve changes to a process, access to the people who run the work today, and a few real examples of the inputs. Stennir also asks who in your company signs off on data and security, because the hosting decision is made inside the audit, not after it.
Where does our data stay if we work with Stennir?
Where your rules require. The hosting decision is written into the scope before anything is built: which country, which cloud and which model can run there. The accounts are in your name and the off switch sits with your team. Stennir works with operations teams in the UAE and Saudi Arabia, so this is a standard first question.
