Explainable AI support for clinicians: Trustworthy AI launches from Évora
A new company in Évora is building an AI decision support system to help healthcare professionals diagnose and prescribe faster-without replacing clinical judgment. Trustworthy AI was founded by Paulo Quaresma and Vítor Nogueira from the University of Évora, and Jianbiao Dai from the University of São José in Macau.
The team is clear on its purpose: assist clinicians, not automate care. The system will offer recommendations alongside clear reasoning and confidence levels so doctors and nurses can decide with more context and speed.
How it works
The AI ingests symptoms, medical history, context and patient characteristics. It then proposes diagnostic and therapeutic options, showing why it reached each suggestion and how confident it is.
It keeps a learning history across many clinical situations. That allows it to flag considerations a clinician might not be weighing in the moment, acting as a second set of eyes.
Why it matters for your workflow
- Time savings: Faster differentials and treatment options during triage, handovers or complex cases.
- Transparent reasoning: Explanations attached to each suggestion so you can validate or dismiss with minimal friction.
- Auditability: Built for external review (with authorization) to trace how outputs were generated.
- Safety net: Alerts for overlooked patterns based on prior similar cases in its learning history.
- Flexible settings: Intended for hospitals, health centres and nursing homes.
Governance and oversight
The system is being built on explainable, auditable and ethical AI methodologies to increase confidence and accountability. That includes the ability to show why each response was produced and to allow outside auditing when permitted.
For context on current guidance, see the EU's ethics guidelines for trustworthy AI and the WHO's report on ethics and governance of AI for health.
Pilots and timeline
Trustworthy AI is based at the Alentejo Science and Technology Park (PACT). The company is preparing applications for EU funding via the Alentejo 2030 regional programme and seeking support in Macau.
Two pilot projects-one in Alentejo and one in Macau-are planned over the next year to test and evaluate the solution. If results meet expectations, broader rollout is targeted by the end of 2026.
What clinicians can expect during pilots
- Decision support, not automation: Final decisions remain with clinicians, with AI providing structured options and rationales.
- Confidence levels: Each suggestion is accompanied by a confidence estimate to calibrate trust.
- Explainability on demand: Quick access to the factors that most influenced a recommendation.
- Continuous improvement: Feedback loops from real cases to refine performance and reduce noise.
What you can do now
- Identify high-friction points where explainable AI could help (e.g., triage, complex differentials, discharge planning).
- Confirm data governance, consent and audit processes for any pilot participation.
- Plan brief training for clinicians on reading AI explanations and managing confidence levels.
- If your organization is preparing to upskill teams on AI oversight, explore job-based learning options: Complete AI Training - courses by job.
Bottom line: Trustworthy AI is pursuing a clinical support tool that explains its reasoning, can be audited and fits into everyday practice. If the pilots validate speed and safety, it could help clinicians deliver decisions faster with clearer accountability.
Your membership also unlocks:
