AI consultancy · Barcelona

AI consulting for companies. From personal AI to process AI.

Your team already uses AI, each person their own way. Together, we map what drains your team, connect AI where it pays off and train your people to keep it running without us.

See where to start

It starts with a Full diagnosis: a 30-min video call with Albert Puig (Co-founder · CPO). Free.

0

penalties if you stop when a block closes

0

changes to your systems until you decide

1

cronuts partner in your mapping sessions

4.9/5 · 40 Google reviews

The essentials

What is AI consulting at cronuts.digital?

AI consulting helps a company decide where to use AI and put it to work in its processes. At cronuts.digital, a Barcelona B2B growth marketing and AI consultancy, it runs in three phases: AI Map, Implementation and Adoption. We map each process with the people who run it, put a value on every opportunity with its hypothesis in writing, and move in fixed-scope blocks, with the option to stop without penalty at the close of each one.

Method

The map comes first. AI comes after, and only where it pays off.

The AI pilots we see fail tend to look alike: the tool comes before the process, no numbers are written down and nobody inside drives them. And if you automate a broken process, you just reach the disaster faster. That's why the AI Map starts with the people who do the work.

  1. With many departments, we start with three

    A project that lands on every department at once creates noise and mistrust. We start with three departments, which act as ambassadors and tell the rest what they've seen, in their own words. You decide when the rest come in.

    You set the order

  2. One session with each process owner

    Forty to sixty minutes, with a short questionnaire or a voice note beforehand: we do the prep, you do the talking. What they do, what adds value and what drains them. Where needed, a second deep-dive session.

    Each process owner and cronuts

  3. Your IT, from the design stage

    If you have an IT team, they confirm what's feasible before anyone gets their hopes up.

    Your IT

  4. You decide here

    First decision

    When the mapping closes, you have the process map and the preliminary opportunity catalogue. You decide what gets a deeper look, or you stop here without penalty.

  5. Business cases and proofs of concept

    Every opportunity gets its business case and one of four verdicts. We prioritise two or three and leave one or two proofs of concept running, on read-only data or on a copy, outside your systems.

    cronuts and your IT

  6. You decide here

    Second decision

    With the roadmap in front of you: estimated effort and budget for each project. You decide what gets implemented, what doesn't and in what order.

See where to start

Who does the work

Your AI Map, led by a partner.

One of these partners leads your AI Map and takes part in the sessions with your process owners. Behind them, the cronuts team builds each project.

  • Albert Puig

    Co-founder · CPO

    Leads product and applied AI, and teaches AI and Growth Marketing at ISDI and IED Barcelona. He's the one you meet in the Full diagnosis and, when he leads your AI Map, he takes part in the sessions with your process owners.

  • Borja Planells

    CEO and Founding Partner

    Specialist in B2B growth marketing and digital growth systems. When he leads your AI Map, he takes part in the sessions with your process owners and presents the roadmap to your leadership team.

    View profile · Borja Planells

Scope

Every opportunity gets its own business case. With the hypothesis written next to it.

It's the document you decide with: if the hypothesis behind an opportunity doesn't hold up, you see it in the business case before you invest in implementing it. The real business case also covers saving or revenue, ROI, complexity, time to implement and dependencies. And if an opportunity screens candidates or evaluates or monitors people, the business case flags it as high-risk, and we don't build that one.

Business case · Preparing custom quotes

Financial impact
High. With the hypothesis written alongside: what it costs today to put each quote together by hand, times the number that come in each month.
Investment
Medium, licences included.
Payback
In months, from the investment and the saving in this same business case.
User buy-in
High. The person who prepares the quotes wants to stop putting them together by hand. Without that, nobody adopts it.
Risk if it fails
Low. A person reviews every quote before it goes out.
Verdict
With AI: there is a return, data and someone who wants to use it. The other possible verdicts are “without AI”, “fix the process first” and “not worth it”.Another example: matching delivery notes against invoices comes out as “without AI”, because a rule in the ERP handles it. A process that doesn't need AI is a result too: it saves you the investment.
An illustration of the format. The examples are made up and are not any client's data: the real values come from your processes and your data.

Phases and how to engage us

Fixed-scope blocks. You can stop at the end of each one.

Each block has fixed scope, deliverables and price, and you know them before work starts. When it closes, you decide whether to continue, with the deliverables in hand.

  1. Start here

    AI Map

    In two blocks. First, the mapping: sessions with your process owners, a process map and a preliminary opportunity catalogue. Then the business cases, the proofs of concept and the roadmap. The timeline goes in the proposal, because it depends on how many departments take part.

    Fixed-scope proposal after the Full diagnosis.

    What cronuts does

    • Puts a partner in charge, Albert Puig or Borja Planells, who takes part in the sessions with your process owners
    • Writes the business cases and builds the proofs of concept
    • Presents the roadmap to your leadership team
    • Sends you a one-page report every two weeks: what's done, what's next and which decision is pending

    What your team does

    • Names one person who can get things moving, even without AI knowledge
    • Talks through their day-to-day in the sessions
    • Grants read-only access, with your IT involved in the design

  2. Implementation

    AI agents and automations for whatever you choose from the roadmap, connected to what you already use: ERP, CRM, email or documents. In four-week sprints.

    A fixed quote per project, with the roadmap in front of you.

    What cronuts does

    • Builds and connects each project
    • Documents every automation

    What your team does

    • Signs off at the end of each sprint
    • Assigns an owner to every automation

  3. Adoption

    Training built on each team's own cases, plus monthly sessions, with the aim of your team running it without us.

    In fixed-scope blocks, like the rest.

    What cronuts does

    • Trains each team on its own cases
    • Reviews every month what gets used and what doesn't

    What your team does

    • Uses and maintains what's been built

In every phase, nothing is written to your systems until you decide.

We don't publish prices, because they depend on how many departments take part and what needs connecting. You get them in writing in the proposal, before work starts. AI licences and APIs for whatever gets implemented are in your company's name and paid at cost. Proposals exclude VAT.

  • Today you only decide on the mapping. The next block is decided when it closes, with the process map and the catalogue in front of you.
  • Two decision points, in writing. When the mapping closes and with the roadmap in front of you. If you don't see a return, you stop without penalty.
  • What has been delivered and paid for belongs to your company, whether you continue or not. If you stop when the mapping closes, you keep the process map and the preliminary opportunity catalogue.

See where to start

Examples

What you can check before you sign.

The first proof is the method: the business case, the verdicts and the two decision points go in writing in your proposal. Below are published cases, each with the phase it demonstrates.

Published case · full-scope project

Antala

.es, .uk and .pt websites integrated with Microsoft Business Central: +20 % revenue and 90 % of new customers via the digital channel.

Implementation connected to the ERP: websites, paid media, nurturing, ABM, Microsoft Business Central integration and a smart product selector. The result belongs to the whole system, not to AI alone.

Read the case

Each case states the phase it demonstrates. No figure in this section is an average or is attributed to AI alone.

See the published cases

4.9/5 · 40 Google reviews

Behind every case, the cronuts.digital team, a B2B growth marketing and AI consultancy.

Is it a fit?

Who it's for. And who it's not for.

Your team already uses AI. Your company doesn't, yet. This works when AI is already on everyone's desk and still has to make it into your processes.

It's a fit if

  • You have several departments and processes that pass from one to another.
  • Your team already uses AI and nobody coordinates it across departments.
  • You can name someone who gets things moving. They don't need to know about AI.
  • You want to decide with numbers and be able to stop if you don't see a return.

It's not a fit if

  • You already know which tool you want and only need someone to install it.
  • You need a saving fixed by contract. We don't promise one. We estimate it in each business case, with its hypothesis, and you test one or two opportunities before implementing.
  • What you want reviewed is your website, for Google and for AI answers. Here we work on your processes.

Service rules

What we never do.

These are the rules we work by, and they go in writing in your proposal.

  1. Write to your systems before you decide

    In the AI Map we work with read-only access. Nothing is written until you decide.

  2. Put AI where it isn't needed

    If a rule, an integration or a tidied-up process solves it, the business case says so and we don't add AI.

  3. Give you a number without its hypothesis

    Every financial impact comes with the hypothesis behind it in writing, so you can challenge it.

  4. Lock you into a tool

    Copilot, Claude, OpenAI or whatever you already have: we choose what integrates best with your systems. Licences for whatever gets implemented are in your company's name.

  5. Leave you with a black box

    Every automation is documented and has a named owner inside your company.

Full diagnosis · Free

Tell us what drains your team. We'll start there.

Write down the process that eats up the most hours on your team and pick a slot for the Full diagnosis, a 30-min video call with Albert Puig (Co-founder · CPO). We look at where AI makes sense in your company, which process the AI Map would start with and what scope fits. If it's not a fit, we'll tell you.

  • Free, with no commitment.
  • Whatever you tell us is confidential. If you need it, we sign a non-disclosure agreement before the Full diagnosis.
  • If it's a fit, you get a fixed-scope proposal for the AI Map.
  • You pick the slot yourself, right after sending.
  • No slot that works? Email us at info@cronuts.digital.

Start here

5 fields, then you pick a slot with Albert Puig.

One sentence is enough. We use it to prepare the video call.

So we can let you know if the video call needs to move.

CRONUTS DIGITAL SL processes your data to handle this request and to measure our campaigns. You can exercise your rights at info@cronuts.digital.

Once you send, Albert Puig's calendar opens so you can pick a slot.

FAQ

What people ask us before getting started.

How much does it cost to implement AI in a business?
It depends on how many departments take part and what needs connecting, so we don't publish a figure that wouldn't hold. After the Full diagnosis you get a fixed-scope proposal for the AI Map, with the scope, timeline and price of each block. Implementation is quoted project by project, with the roadmap in front of you. Licences and APIs for whatever gets implemented are at cost and in your company's name. Proposals exclude VAT.
How much does an AI consultant charge?
We don't publish rates, because each project depends on how many departments take part and what needs connecting. You don't sign blind either: every block has a fixed scope and price, and you get both in writing in the proposal before we start. Implementation is quoted project by project, with the roadmap in front of you. The first step is the Full diagnosis. Free, with no commitment.
What happens if I want to stop?
You can stop at the close of any block, with no penalty. If you stop partway through a block, you pay the proportional share of the block in progress and keep what has been delivered up to that day. What has been delivered and paid for belongs to your company, whether you continue or not. If you stop when the mapping closes, you keep the process map and the preliminary opportunity catalogue.
How long does the AI Map take?
The duration is written into your proposal, because it depends on how many departments take part and on your process owners' calendars. Each of them gives one session of forty to sixty minutes, and a second one where needed. With many departments, we start with three. When the mapping closes, you decide whether to continue.
My team already uses AI on its own. Why would I need a consultancy?
That's personal AI: each person in their own browser tab, whenever they remember. It saves one person a few minutes. Process AI changes how a whole department works: it's connected to your data, it has an owner and it sits where the numbers say it pays off. We make that leap with you, so you can keep making it without us. The licences you already pay for can be the starting point.
How much time does it take from my team, and how do you get them to use it?
One session of forty to sixty minutes per process owner, a second one where needed, and one person inside who keeps the agenda moving. Before each session, a short questionnaire or a voice note: we do the prep, your team does the talking. To get it used, adoption is a phase of the project. The business case measures the buy-in of whoever will use it, we design it with that person to take away what drains them, and then we train the team with monthly sessions. What you do with the hours it frees up is up to you.
What about my data, the GDPR and the EU AI Act?
In the AI Map we work with read-only access and write nothing to your systems until you decide. We go through how your data is handled with you before work starts. We don't design practices the EU AI Act prohibits. Nor do we build systems that screen candidates, make decisions about working conditions, or evaluate or monitor the people on your team, which that regulation treats as high-risk. If the AI Map finds an opportunity like that, we flag it in its business case and recommend you review it with a specialist law firm. We're not a law firm: if an opportunity needs a legal opinion, we tell you.
What if my ERP is old or my IT team won't open up its systems?
We start from what you already have: ERP, CRM, email, documents or your own platform. If you have an IT team, they're part of the design from the start, so what we propose already takes into account what can be built. If they can't open up APIs or data, we deliver the strategy and an implementation guide for your engineers.
What's the difference between the Full diagnosis and the AI Map?
The Full diagnosis is a first video call to see whether the AI Map fits and where it would start. Free, with no commitment. The AI Map is the project: sessions with your process owners, business cases, proofs of concept and a roadmap, under a fixed-scope proposal.
What does an AI consultant do?
They help you decide where to use artificial intelligence in your company, and then put it to work. At cronuts, that starts with understanding each process with the person who runs it, and the tool comes after. Then we connect AI to your systems where it pays off and train your team to use it. Sometimes the verdict is that a process doesn't need AI, and we tell you that too.
How do you choose an AI consultancy?
Before you sign, ask for it in writing: how they decide where AI goes, which hypothesis backs each number, when you can stop and what you keep if you do. Ask who will sit in the sessions with your process owners, and whether they'll tell you when a process doesn't need AI. If the proposal starts with a tool instead of your processes, be wary. At cronuts all of that goes in the proposal: a partner leads the AI Map, every opportunity gets its business case and you can stop at the close of each block.
How do I implement AI in my business?
In order. First, map what each department does and what drains it. Then put numbers on each opportunity, with its hypothesis, and pick the ones with the best return. Test one or two before investing more. Only then build, in short sprints, with someone on your team owning each automation. That's the order we follow from the AI Map to Implementation.
Do you build AI agents and automations?
Yes, when the business case says they pay off. In Implementation we build whatever you choose from the roadmap, whether that's an AI agent, an AI automation or, if it's enough, a rule without AI, and connect it to your ERP, CRM, email or documents. Every automation is documented and has an owner inside your company.
What is the best AI for businesses?
The one that integrates best with what you already use. If your company runs on Microsoft, Copilot is in the comparison from the start. If not, we weigh others, such as Claude or OpenAI, or an automation with no AI at all. We choose on integration, cost and risk, never on hype.

To go deeper, see the published cases, each with the phase it demonstrates.