R&D
Applied research as method, not decoration.
R&D runs through every project — operating or in development. The model is simple to state and demanding to keep: research, build, validate, operate, learn.
The fundamental rule
Development is not validation.
Developing is building: iterate, adjust, improve with continuous feedback. Validating is testing: criteria fixed in advance, documented conditions, a real possibility of failure. When the two are mixed, the process that should test starts selecting success cases — a tool “evaluated” only where it worked was not evaluated, it was demonstrated.
Maturity model
Visible states, added responsibility.
Real operation, with users and service responsibility. Examples: PecasAuto.pt, Byblos Native, Scientia Humanae, IDEG.
Operating and scaling: real data feeds research. Example: the PecasAuto.pt matching and sourcing agenda.
In development and validation, with declared limits. Examples: Cognitive Assessment Platform, Human Cybersecurity, Digital Health & Ageing, Quality & Service Evaluation, AI-assisted scientific publishing.
Definition and prototyping. Example: Assisted Living AI Modules. Concept-stage projects are not published.
R&D lines
Where we are researching.
Cognitive Assessment Platform
Time-based data and intra-individual variability in cognitive assessment. Case →
Assisted Living AI Modules
Modular AI for assisted living, privacy by design. Line →
Human Cybersecurity
Psychology of cybersecurity and social engineering. Line →
Digital Health & Ageing
Autonomy, adherence and digital exclusion. Line →
Quality & Service Evaluation
Indicators and evaluation of social and health services.
AI-assisted scientific publishing
Support for authors, editors and reviewers, with public ethical limits. Module →
Circular Economy & Smart Reuse
Matching and data for reuse markets. Case →
Consumer Psychology & E-commerce
Applied experimentation on trust and purchase decisions.
AI ethics