Back to homeAI consulting and implementation for SMEs

We implement artificial intelligence in Spanish SMEsprivate and compliant with the EU AI Act and GDPR

Tecnea is a Spanish company that designs, builds and implements custom artificial intelligence systems in SMEs: assistants over internal documentation, process automation and AI applications integrated with the ERP and CRM. Data is processed in private environments inside the European Union and every project is documented in line with the EU AI Act and GDPR. Headquartered in Madrid, with an office in Gijón and projects across Spain since 2016.

Over 10 years with Spanish companies
Your data never leaves your environment
Documented AI Act and GDPR compliance

What we deliver

What your company receives in an AI project with Tecnea

An AI project is not a report or a demo. These are the eight deliverables we hold our own work to:

  1. 1. Inventory of the AI tools your company already uses

    Including the ones your team uses on its own: in companies of 50 to 500 people, 64% of employees use unauthorised AI, according to WatchGuard's 2026 report. Without an inventory there is no control and no compliance.

  2. 2. Risk classification of each system

    Under the EU AI Act: prohibited, high risk, transparency or minimal risk. Most SME uses fall outside high risk, but it has to be documented.

  3. 3. Applicable obligations and how they are met

    Staff AI literacy (Article 4), transparency towards anyone interacting with an AI (Article 50), logging and human oversight. Translated into concrete measures, not legal citations.

  4. 4. Review of personal data processing

    Legal basis, minimisation, where data is processed, processors and an impact assessment where required, following the criteria the Spanish data protection authority (AEPD) published in July 2026 on data quality and accuracy in AI systems.

  5. 5. Internal AI usage policy

    Which tools your team may use, with which data and under what supervision. One page, not a manual.

  6. 6. Team training

    The AI literacy obligation is met with short sessions adapted to each role, with attendance records.

  7. 7. Register of systems and suppliers

    Which model is used, who hosts it, which contract covers it and who is accountable. It is the first thing an auditor or a large client asks for.

  8. 8. Technical implementation in production

    What separates a consultancy from a project: the solution connected to your systems and in daily use by your team, with support after delivery.

How we work

Four phases, fixed price for each one

  1. 1 to 2 weeks

    1. Diagnosis

    Process map, use cases prioritised by return and risk, savings estimate and proposed architecture. It ends with a decision: go on or stop.

  2. 4 to 8 weeks

    2. Pilot in production

    A real process, with real users, fixed scope and fixed price. If it does not add value, it stops here.

  3. 8 to 12 weeks

    3. Roll-out

    Integration with ERP, CRM and email, team training, compliance documentation and handover. The code and the data belong to your company.

  4. monthly, optional

    4. Evolution

    Maintenance, new use cases and compliance reviews when regulation or models change.

Who it is for

SMEs that lead their sector and have no IT department

We work with profitable, established companies of 20 to 100 people where one person makes the decision. Typical sectors:

law firms, accounting and tax advisers, notaries, consultancies, engineering and architecture firms, clinics, real estate, insurance brokers, manufacturing, logistics and distribution, construction, schools, agencies and retail.

Projects

Examples of what we build

  • Law firm

    Drafting of legal documents and case-law search with pseudonymised health data: a private AI system, with lawyer review of every document, designed as special-category processing under Article 9 GDPR.

  • Wholesale distributor

    B2B ordering portal with a voice and text assistant for customers, integrated with the ERP: prices, stock and terms always come from the management system, never made up by the AI.

  • Construction company

    Site reports from the phone, automatic reading of delivery notes and invoices, and cost control against each project's budget.

Where

Madrid, Gijón and all of Spain

Headquarters in Madrid (Velázquez 27) and an office in Gijón, at the Science and Technology Park. Most of the work is done remotely, with in-person meetings in Madrid and Asturias when the project calls for it. We have clients in Madrid, Asturias and the Balearic Islands, among other regions.

Frequently asked questions

What we get asked before starting

Do I need an IT department to implement AI in my SME?+

No. Tecnea designs, builds and maintains the system; your company brings the business knowledge and one point of contact. Most of our clients have no in-house technical staff.

Where is my company's data processed?+

On infrastructure inside the European Union, and in your own environment when the sensitivity of the data requires it. None of your data is used to train third-party models, and this is written into the contract.

Which EU AI Act obligations apply to an SME in 2026?+

Since 2 August 2026: training the staff who use AI (Article 4), informing people when they interact with an AI system and labelling certain generated content (Article 50), and not using prohibited practices. High-risk obligations were postponed to 2 December 2027 and affect specific uses, such as recruitment.

How much does an AI project cost for an SME?+

It depends on the process and the integrations. We fix the price per phase, starting with a short diagnosis and a limited-scope pilot, so the investment only grows once the return is proven. The calculator on this site estimates the savings for your sector in one minute.

How long does it take to be up and running?+

A pilot in production takes four to eight weeks from the diagnosis. A full roll-out, with integrations and training, eight to twelve.

My team already uses ChatGPT on its own. Is that a problem?+

It is common and it is a risk: client data in uncontrolled tools with no traceability. The first deliverable of the project is precisely that inventory, and the fix is usually to give them a private tool that is better than the one they use quietly.

Which of your company's processes can be automated with AI?

Fill in the one-minute questionnaire and receive a report with the use cases for your sector and a savings estimate. Or write to us and we will look at it together.