Ideas Medioambientales wins an Ibermutua award for PrevIA, its AI-based occupational safety assistant

Ibermutua has announced the jury's decision for the 4th edition of its “Eladio González Malmierca” Occupational Risk Prevention Awards, and Ideas Medioambientales, S.L. has been awarded in the category for innovation and implementation of occupational risk prevention and health technologies for PrevIA, an artificial intelligence-based occupational health and safety (OHS) assistant.

PrevIA is an in-house tool that puts the company's safety protocols on every field worker's phone, at any time and in plain language. This post summarises what the award recognises, what the tool does and how we built it.

What are the Ibermutua “Eladio González Malmierca” Awards?

They are the awards through which Ibermutua recognises good occupational health and safety practices among its member companies, as well as initiatives by media outlets, institutions and professionals. They honour Eladio González Malmierca, for many years Ibermutua's Director of Prevention and a reference in the field in Spain and internationally.

The 2026 edition coincides with the mutual's centenary. The jury included representatives of the Regional Institute for Occupational Health and Safety of the Community of Madrid, the Regional Institute for Occupational Health and Safety of Castilla-La Mancha and the BeOn network, together with Ibermutua's Institutional Relations and Prevention departments and representatives of member companies awarded in previous editions.

The award ceremony will take place on Tuesday 17 November 2026 at 9:00 at Ibermutua's head office in Madrid.

What is PrevIA and what problem does it solve?

Most of our work takes place in the field: forests, riverbanks, forest plots, linear infrastructure and protected areas. Our teams face potentially dangerous wildlife, work at height, adverse weather, prolonged heat exposure and, often, lone working with patchy coverage.

Until now, protocols were PDF documents on the corporate server, available only from the office, and any urgent question depended on a phone call to the OHS manager. PrevIA reverses that model: it is a web application that opens from the phone's browser, with nothing to install, and brings preventive knowledge to the place where it is needed.

What features does PrevIA have?

• OHS chat with AI. Workers ask in natural language (“a snake has bitten me, what do I do?”) and get a practical answer. The system includes 12 quick-query buttons for the most frequent risks in the sector.
• Weather panel with alerts. Using the device's geolocation, it shows temperature, wind, humidity and UV index, and triggers alerts tied to the thresholds defined in the company's protocols.
• Emergency button for lone workers. It emails the OHS manager the GPS coordinates with a Google Maps link. If there is no coverage, it stores the signal on the device and retries every 30 seconds.
• Incident reporting with photo. Workers describe the incident, give the location and attach a photo from the phone, in real time.
• AI-generated quizzes. From the protocols, the OHS manager generates multiple-choice questions, reviews and edits them before sending them to staff.
• Admin panel. The OHS technician manages protocols and quizzes without touching code.

How does the AI answer without making things up?

PrevIA is not a generic chatbot connected to the internet. Each query is answered using the protocols the company itself has loaded into its private repository, and the AI interprets and explains them in plain language. If a protocol changes, it is enough to upload the new document from the panel: there is no retraining. And if the answer is not in the loaded documentation, the system says so instead of improvising.

The tool does not replace the OHS technician: it multiplies their reach. The technician defines the protocols, reviews incidents, manages quizzes and administers the system. The AI always works within the framework the technician has set.

How was it built and why is it innovative?

The conversational assistant relies on the Claude language model from Anthropic, configured with our protocols. The rest of the system was developed with free or pay-per-use platforms, and the only variable cost is the model API usage.

According to the report submitted to the jury, the project was developed in under three weeks and without prior programming knowledge, through AI-assisted development: describing the preventive problem in natural language and letting the AI generate the code, with human review. The lesson we want to share is that, in this approach, the key skill shifts from programming to defining the risk and the protocol well.

What comes next

A worker-tracking panel is under development and integration with the Microsoft 365 ecosystem (Teams, SharePoint, Power Automate and Power Apps) is planned as the next phase. The methodology is replicable by any company: identify the protocols, create the platform accounts, generate the code with AI and load your own protocols from the panel.

We thank Ibermutua and the jury for this recognition, which we share with the team that contributes every day the protocols and field experience PrevIA relies on.

If you want to explore how to apply artificial intelligence to your organisation's safety or document management, get in touch with our team.