Skip to content

AI consulting

AI Consulting: Process First, Technology Second

Together we find where AI will genuinely make a difference in your business, prioritise it, and turn it into a working setup your team can start using the next day.

Two people mapping a process with sticky notes
On this page
  1. What is AI consulting?
  2. What AI consulting is not
  3. Who is it for?
  4. How I work
  5. How does it fit your team and existing systems?
  6. Which work can AI genuinely take over?
  7. Related services
  8. Why I do this work
  9. Frequently asked questions
  10. Related articles
  11. How we start

What is AI consulting?

AI consulting is the work of examining how a business currently operates, identifying where AI will make a real difference, prioritising that point and turning it into a working setup. It is not selling tools, and it is certainly not producing slide decks. The output is a system your team can start using the next day.

A great deal has been said about AI over the last two years, yet most businesses are still where they started. The reason isn’t a lack of tools; there are more than enough. The real problem is that nobody looks at how the work actually runs first. Starting with a tool and trying to work back to the process is why so many projects are quietly abandoned a few months later.

I work the other way round. First I try to understand how things run today: where time is lost, which tasks are done by hand over and over, which processes are more complicated than they need to be. Until that is clear, there is no point discussing which tool to use.

What AI consulting is not

Making this distinction early prevents disappointment later.

  • It is not a tool demo. A presentation of the latest models gains you nothing. What matters is which tool takes over which part of your work.
  • It is not a one-off training session. Training can be part of it, but on its own it rarely changes anything. The team leaves the session and goes back to the old way the following week.
  • It is not a promise to automate everything. Some work is not suited to automation. Forcing it costs more than doing it by hand.
  • It is not a setup that keeps you dependent on me. Everything I build is documented and handed over to your team. Ongoing support is optional, not a requirement.

Who is it for?

I work with three kinds of organisation. If you recognise yourself in one of them, the discovery call will be useful.

ProfileTypical situationUsual first step
Small and mid-sized businessA small team doing a lot; everyone has their own method; repetitive work is on nobody’s agendaA process map and a single pilot automation
Agency or service firmAs the client list grows, reporting, content and follow-up work multipliesAutomating reporting and content production workflows
Corporate team or departmentTools are available but used inconsistently; no clear policy on data securityA usage policy, team training and standard workflows

On the other hand, for a very new business whose processes haven’t settled yet, AI is usually not the right first investment. You can’t automate a process that doesn’t exist. If that’s your situation, I will say so in the call.

How I work

I follow a four-step process. Each step ends with something concrete in your hands, so you can see what you are getting before moving to the next.

StepWhat happensWhat you have afterwards
1. DiscoveryWe talk through how things run: who does what and how, how often each task repeats, where work waitsProcess map and a list of time drains
2. PrioritisationWe rank the opportunities by impact and difficulty. Not everything is done at oncePrioritised roadmap and a chosen pilot
3. ImplementationThe chosen area is built, tested and run with real workA working automation or workflow
4. Handover and supportTraining for the team, documentation, what to do if something goes wrongDocumentation, training and optional maintenance

The most important step is the second one. Most projects fail not here but because they skip it. Any business has dozens of things that could be automated; trying to do them all at once means doing none of them well. I explain how I decide whether a task suits automation in the five-question filter article.

How does it fit your team and existing systems?

The setup sits on top of the software you already use; I don’t require you to move to a new system. During discovery I list the tools you use and check which of them allow outside connections. If something can’t be connected, I tell you at the start.

The person who does the work by hand today is involved from the beginning, because they know the process best and they are the one who will adopt the new way of working. At handover I train the team, leave written documentation and explain what to check first if something goes wrong.

Which work can AI genuinely take over?

These are the areas I meet most often in small and mid-sized businesses, and the ones that tend to deliver results fastest. The list isn’t exhaustive; it’s here to give you a starting point.

AreaTypical repetitive taskFit for automation
Customer communicationAnswering the same questions repeatedly; slow first responsesHigh
ReportingCollecting data from several sources and turning it into a report by handVery high
Content productionDrafting, reformatting, adapting for different platformsHigh
Data entry and transferCopying from one system to another by handVery high
Proposals and documentsFilling in the same template by hand every timeHigh
Research and summarisingReading long documents and writing summariesMedium
Work that needs judgementInterpretation, negotiation, managing exceptionsLow; keep it with people

The last row is there on purpose. The fact that a task can be automated doesn’t mean it should be. Work that needs judgement, shapes the relationship with a customer or carries a high cost of error should stay with people. Part of consulting is deciding what you won’t automate.

Assess one of your own tasks: the automation calculator shows in five questions whether a task is a good candidate for automation and how much time it could free up each week.

After the discovery call, work usually develops into one or more of these.

Why I do this work

Not everyone offering AI consulting has a technical background. I do, and it shapes how I look at the work.

I studied statistics at Ondokuz Mayıs University, completed an MSc in Data Science at the University of Southampton, and worked as a data engineer at BP for part of my career. That background taught me to ask less about what the model is and more about where the data comes from, how a process breaks and why a system isn’t working. The rest of the story is on the About page.

I have also been running my own content operation for three years and built my channels from scratch. That teaches something consulting theory doesn’t: building a system is easy; keeping it running every week is hard. I design automations with that in mind.

Frequently asked questions

How long does AI consulting take?

The discovery call takes 20 minutes. Process analysis usually takes one to two weeks. Getting a pilot automation up and running typically takes two to four weeks. Larger scopes take longer, but I don’t want you to make a long commitment before seeing a first concrete result.

We don’t have a technical team. Can we still work together?

Yes. I do the setup; all your team needs to do is use the tool that comes out of it. At handover I provide training and leave written documentation.

Will our data be safe?

Data security is discussed on the first day of a project, not added later. We decide together which data goes to which tool, we leave sensitive data out of processes that don’t need it, and where necessary we consider setups in which data never leaves your organisation. If you have legal obligations, we need to discuss them before the build.

How much does AI consulting cost?

I don’t publish a fixed price list, because no two businesses need the same thing. The main factors are how many processes are in scope, how complex your current systems are, whether custom development is needed and whether you want support after handover. After the discovery call I send the scope and price in writing. I go through the cost factors in detail in a separate article.

Can we manage the system ourselves afterwards?

Yes, that’s the aim. Everything I build is documented and accounts are opened in your name. I don’t build anything that leaves you dependent on me. If you want ongoing support we can arrange it; if not, the system stays with you and keeps running.

Which tools do you use?

Tools are chosen after the process is clear, not before. Depending on the need, that might be automation platforms, language models, your existing software’s own integrations or custom code where necessary. I’m not tied to any particular tool, because then the solution would be shaped by the tool rather than the need.

How we start

The starting point is always the same: a free 20-minute discovery call. I don’t try to sell you anything in that call. We talk about how the work runs, where it gets stuck and what is worth automating.

The call ends in one of two ways. Either we find an area where working together makes sense and I send you the scope in writing, or I tell you that you don’t need me for this. The second happens often, and it’s better than wasting anyone’s time.

20 minutes · free · online

Choose a time that suits you

  • Times are shown in your own time zone (UK or Türkiye)
  • No preparation needed; one recurring task is enough
  • In Turkish or English

or send an email: fahriaksoy.tr@gmail.com

Request a discovery call

Tell me briefly what you would like to discuss and which time suits you. I will reply within one working day with a confirmed time.