Artificial intelligence & machine learning
Investigate a learning problem, develop or adapt a method, and assess it against the task it is meant to solve.
Classification, anomaly detection, model calibration, and course-specific AI.
Scientific consultancy
Research-led expertise for problems that need more than a standard answer.
Machine learning, artificial intelligence, data science, image processing, and statistical analysis—connected by mathematical reasoning and practical software development.
Discuss your projectAreas of expertise
These are starting points for a conversation. An engagement may combine several disciplines; its feasibility, deliverables, and terms are agreed individually.
Investigate a learning problem, develop or adapt a method, and assess it against the task it is meant to solve.
Classification, anomaly detection, model calibration, and course-specific AI.
Make complex data useful: explore patterns, test hypotheses, compare approaches, and communicate the findings.
Exploratory analysis, predictive modelling, data preparation, and interpretable reporting.
Design an analysis that accounts for uncertainty, limited samples, and the decisions the results will inform.
Study design, ROC analysis, cross-validation, uncertainty estimation, and model comparison.
Develop and evaluate methods that extract useful information from images, with the application’s constraints in view.
Image enhancement, feature extraction, pattern recognition, and medical-imaging research.
Give analysts useful ways to investigate the same observations across views and connect visual evidence to computation.
Linked plots, anomaly exploration, scientific interfaces, and analytical workflows.
Translate a method into working, documented software, with tests and a handover suited to your team.
Research prototypes, algorithm implementation, workflow integration, and reproducible tools.
Research into practice
Two examples of our work connecting research methods with practical software in image analysis and anomaly detection.
Image analysis
Image-processing research translated into software for mammography. LIBCAD and WEBCAD bring image enhancement and abnormality-detection methods into research workflows.
Explore CAD U.S. Patent 10,789,712 B2 — Image analysis to detect cancerAnomaly detection & visualization
UN-AVOIDS combines unsupervised anomaly scoring with a visual approach to investigating outliers. VAAD brings this work into linked views that help analysts explore unusual observations.
Explore VAAD U.S. Patent Application Publication 2022/0327182 A1 — UN-AVOIDSWork with MESC Labs
Get research-led advice, analysis, and custom implementation. Share your objective, available data, and timeframe so we can discuss an approach, scope, and quotation.
Please do not email confidential datasets before we agree a secure transfer arrangement.