Explainable AI
“Why did the model produce this output?”
Attribution and interpretability methods that reveal what a model relies on — and checks that those explanations are themselves reliable.
Research and innovation
We treat AI as a scientific practice: state the question, test against honest baselines, explain what the model learned and check it holds on new data. That discipline carries straight into the products we build.
Areas of focus
Stability and robustness
“Does the result hold on new data?”
Evaluating whether models and their explanations stay consistent across datasets, sites and recording conditions.
Responsible AI
“Who could this system fail?”
Subgroup analysis, bias assessment and governance practices built into development rather than audited afterwards.
Healthcare and biomedical AI
“Is this clinically meaningful?”
Applying machine learning to biosignals, clinical records and biomedical data with domain experts in the loop.
Human-centred AI
“Will people use it well?”
Designing how outputs and explanations are presented so people neither over-trust nor ignore them.
AI evaluation
“How good is good enough?”
Evaluation protocols matched to real use: the right metrics, honest baselines and reporting of failure cases.
Publications and research outputs
Papers, preprints, datasets and open-source tools will be listed here with full citations.
[CONTENT REQUIRED]List verified publications (authors, title, venue, year, link)
Only peer-reviewed or publicly available work attributable to NovaRqis or its team should be listed.
Collaborate with us
We work with universities, research groups and clinical researchers on applied AI questions.