App Development

AI by and for people: lessons learned from 1,000 hours of training at GooApps®

In 2025, GooApps® carried out an internal transformation plan with more than 1,000 hours invested in R&D and training. The goal was not simply to “adopt tools,” but to establish a solid ethical and technical criterion. We structured our knowledge across three levels (basic, intermediate, and advanced) to ensure that AI is applied responsibly—improving speed without compromising security or end-user privacy.

Why did we choose training instead of just “using” AI?

Generative AI is accessible, but it can be dangerous if used without understanding. At GooApps®, we recognized early on that efficiency cannot come at the expense of control. Our motto throughout 2025 was clear: “AI is not magic; if we use it, we must be able to explain it, test it, and take responsibility for its outcomes.”

The problem we address with this training is not technological, it is one of trust. A health or wellness product cannot afford hallucinations or undetected biases.

To uphold this principle in daily practice, at GooApps® we translate it into a clear methodology for the responsible use of AI, with technical criteria, human validation, and context control, allowing us to scale without losing reliability or traceability.

The 3 competency levels: structure of the 2025 plan

To scale knowledge in an orderly way, we designed a cross-functional plan divided into three layers of depth. This ensures that everyone—from management to software architects—speaks the same language, while using tools adapted to their role.

Competency matrix by level

Ethics and responsibility: the validation checklist

We do not implement AI solutions “just because.” To ensure that technology delivers real value and respects people, we established a mandatory validation checklist before starting any development involving intelligent components:

1. Human value: what real human problem are we solving?
2. Privacy: what data is involved and what are the exposure risks?
3. Failure analysis: what happens if the model is wrong or hallucinates?
4. Validation (QA): how will we test and measure accuracy?
5. User control: does the user understand they are interacting with AI, and can they take control?
6. Explainability: can we explain why the AI made that decision?

If we cannot satisfactorily answer these questions, the feature is not developed.

Real impact for the client: what actually changes in projects?

This internal training translates into tangible benefits for our clients and for the users of their applications. We do not sell “AI”—we sell results enhanced by AI.

• More precise definitions: Product Managers use AI to challenge requirements and reduce ambiguity before development begins.
• Speed with traceability: we automate repetitive tasks (boilerplate, unit tests) while keeping human control over architecture.
• Predictive quality: our QA teams use AI to detect edge cases that humans might miss due to fatigue.
• More human experiences: in health and sports, we use AI to personalize plans and content, moving away from one-size-fits-all solutions.

What’s next in 2026: continuity and responsibility

If 2025 was the year of learning and standardization, 2026 will be the year of deep integration. Our focus remains on building technology that is useful, secure, and human.

AI will change how we work, but at GooApps® we are clear about what will never change: why we work—to improve people’s lives through health, sports, and wellbeing.