# AI consulting services

> AI consulting services that go all the way to implementation: we pick what to automate, build the pilot and scale it with you. Free first call.

URL: https://gradual.pro/en/services/ai-consulting

Updated: 2026-09-29

We decide with you what to automate first, then build it. Strategy and implementation from the same team.

- Start with a free 30-minute call; after it we send a written proposal with a fixed budget per phase.
- We handle strategy and implementation: we decide what to automate, build it and leave it running.
- Projects move in phases: diagnosis, pilot, scale and operation. Every phase has a written savings target.
- Diagnosis is the first phase of the project. It takes two to three weeks and exists to choose the pilot.

## What do AI consulting services include?

AI consulting services analyze how your company works, decide where AI saves real time or money and, at Gradual, implement it. The output is not a document. It is a process that someone used to do by hand and now runs on its own, with the hours saved measured.

In a business of 10 to 100 people the same candidates show up almost every time: emails someone sorts by hand, invoices keyed into accounting software, quotes assembled by copying data from three places, calls nobody picks up after hours. We put a number on each one (hours per month, errors, lost customers), pick the one that pays back fastest and build it.

If a process does not justify the investment, we tell you and leave it alone.

## Why combine consulting and implementation?

Because most AI projects stall between the recommendation and the rollout. A consultancy hands over a report, a different vendor reads it their own way, and nobody owns the result.

When the same team decides and builds, recommendations are realistic from day one: we only propose what we know how to build, with the real cost of building and running it. And if the pilot misses the agreed savings target, that is our problem, not a third party's.

- One point of contact from the first call through to operation.
- Cost estimates based on what we will build ourselves.
- Clear accountability for the outcome of every phase.

## How does the phased approach work?

AI implementation goes wrong when a company tries to change everything at once. So we work in four phases, and we do not start the next one until the previous one has hit its target. You can read the detail on [our method](/en/method).

- **Diagnosis** (2-3 weeks): interviews, a review of your tools and task measurement to choose the pilot and set its savings target.
- **Pilot**: we build the first [business process automation](/en/services/ai-automation) or [AI agent](/en/services/ai-agent-development) and put it into production with real data.
- **Scale**: if the pilot delivers, we extend the solution to more processes, teams or channels.
- **Operation**: we maintain, measure and tune what already works, with a monthly results report.

## What do we decide together in the strategy phase?

The strategy work is short and concrete. We skip the trend talk and focus on which process to automate first, with which tool and against which measurable target.

For each candidate we look at the hours it takes today, the kind of solution that fits (an automation, an agent, an integration between tools or [custom software](/en/services/custom-software-development)), the cost to build it, the monthly running cost and the risks: personal data, vendor lock-in, or processes that need tidying up before they can be automated. From there we choose the pilot and put it in writing before we start building.

## When does it make sense to hire an AI consultant?

It makes sense when you know your team loses hours to repetitive tasks but you do not know where to start, or when you already tried an AI tool on your own and it never got past the trial. It also helps when several vendors have pitched you different things and you want someone to choose based on your numbers and then make it work.

If you are not sure where your company stands, take the free [AI readiness assessment](/en/tools/ai-readiness-assessment) first. It takes five minutes and lets you arrive at the call with context.

## Process

1. **Free intro call**: 30 minutes to understand your business. A written proposal with a fixed budget per phase follows.
2. **Diagnosis**: In one to two weeks we measure processes, choose the pilot and set its savings target.
3. **Pilot in production**: We build the first automation or agent and put it to work with real data.
4. **Scale**: If the pilot hits its target, we extend the solution to more processes or teams.
5. **Operation and improvement**: We maintain what works, measure the savings and send you a report every month.

- typical length of the diagnosis phase within a project: 2-3 weeks
- typical time to get the pilot into production: 4-6 weeks
- typical payback period for a well-chosen pilot: 3-6 months

Indicative ranges common in companies of 5 to 200 people. They vary with the process chosen, the integrations needed and the quality of the data available.

### Traditional consultancy (report) vs Gradual (consulting and implementation)

|  | Traditional consultancy | Gradual |

| --- | --- | --- |

| What you get | Long strategy report | Automated processes up and running |

| Who implements it | Usually another vendor | The same team that recommends it |

| Time to results | Months of analysis before anything is built | Pilot in production within weeks |

| Written savings target | Sometimes | Yes, in every phase |

| Budget | Billed by consulting hours | Fixed per phase before we start |

| Entry point | Paid analysis project | Free 30-minute call |

Tools: OpenAI, Anthropic Claude, Google Gemini, n8n, Make, Microsoft 365, Google Workspace, HubSpot, QuickBooks, Xero

## FAQ

### How much do AI consulting services cost?

It depends on the process chosen for the pilot, the number of integrations, the data volume and how many phases you want to cover. The first 30-minute call is free. After it we send a written proposal with a fixed budget per phase, agreed before we start. Third-party usage (AI APIs, telephony, hosting) is paid directly to the providers. [Tell us about your case](/en/contact).

### What is the difference between AI consulting and implementation?

Consulting decides what to do and in what order. Implementation builds it and puts it live. At Gradual they are part of the same project: diagnosis is the first phase and ends by choosing the pilot we build next, with no change of team.

### Can I hire you for just the diagnosis or a plan?

No. We do not sell standalone reports or plans. Diagnosis is the first phase of an implementation project, and its job is to pick the right pilot. If the intro call shows that AI is not a fit for your company yet, we will tell you, with no obligation.

### Do I need clean data before starting?

No. Part of the diagnosis is finding out where your data lives and what state it is in. If something needs tidying before it can be automated, the proposal includes it with its cost. That cleanup often saves time on its own.

### Is AI consulting worth it for a small business?

Yes, and small businesses often see returns sooner because there are fewer systems to connect. From about five people doing repetitive admin work, there is usually at least one process that pays for itself within a few months.

### Can you work with a company outside Spain?

Yes. Gradual is based in Spain and works with clients in the UK, the US and elsewhere remotely, in English. Calls, workshops and handover happen over video, and we schedule around your time zone.

### How do you handle confidential information during the project?

We sign a non-disclosure agreement before starting and only request the access each phase actually needs. Personal data is handled in line with GDPR, and we never use your information to train AI models.