Projects

Moodle-Based Learning Platform

Need: A healthcare-education project needed a structured, usable learning platform that teams could actually run and maintain.

What I did: Designed, built, configured, and deployed a Moodle-based LMS end to end — server setup, custom plugins for usability and course structure, and the user-facing experience.

Outcome: A full production deployment, live and serving active healthcare-education cohorts.

→ Live production platform serving active healthcare education cohorts

REDCap Deployment on AWS

Need: A research team needed a secure, self-hosted platform to capture and manage study data.

What I did: Provisioned and configured REDCap on AWS (Lightsail) — server and environment setup, application configuration, and user onboarding — selecting the right tool for the requirement rather than building from scratch.

Outcome: A live REDCap instance in active use by research users to collect and manage data.

→ Live REDCap instance deployed on AWS and in active use

Housing Market Forecasting

Built forecasting models for house price and rent analysis in Python — data cleaning, exploratory analysis, feature engineering, statistical modelling, and visualisation — applied to real Irish housing-market data to surface trends and anomalies.

Dublin Bikes Data Pipeline

Built an end-to-end pipeline using the Dublin Bikes public API — automating collection, cleaning, transformation, and visualisation for exploratory analysis and reporting. The same data-flow logic that applies to clinical data.

Clinical Data Transparency Ontology

Problem: Healthcare data lifecycles are opaque — it's often impossible to trace where a data point came from, who acted on it, or why it looks the way it does.

Approach: Built an ontology using PROV-O and SEIPS to formally represent data provenance, actor accountability, and workflow context in clinical environments. Combined data flow diagrams, UML modelling, and data value mapping.

Outcome: A structured, queryable model of how data moves and changes across clinical systems — making the invisible visible.

→ Produced a queryable provenance model making clinical data flows transparent and auditable

Healthcare Work System Ontology

Problem: AI systems in healthcare are often evaluated on technical performance alone, ignoring the work systems they are embedded in — the people, tasks, tools, and environments that determine whether a system succeeds or fails.

Approach: Developed a formal ontology based on the SEIPS (Systems Engineering Initiative for Patient Safety) model to represent healthcare work systems and their socio-technical context around data use and AI adoption.

Outcome: A reusable semantic model that supports systematic analysis of how AI interacts with clinical work — applicable to governance, procurement, and deployment evaluation.

→ Reusable semantic model applicable to AI governance, procurement, and deployment evaluation

Clinical Data Flow Mapping & Governance

Problem: A major acute hospital had fragmented data collection, unclear reporting accountability, and gaps between what was recorded and what was clinically meaningful.

Approach: Analysed data collection and quality-management practices across hospital departments. Mapped data flows, identified governance gaps, and documented reporting standards for clinical and senior operational stakeholders.

Outcome: Structured outputs and workflow maps that surfaced systemic data-quality issues — and informed how similar problems could be addressed at a design level, not just operationally.

→ Outputs and workflow maps used by senior clinical and operational stakeholders

Inclusive Design for Healthcare AI

Problem: AI systems in healthcare are frequently designed without meaningful input from the people who will use them — leading to poor adoption, workarounds, and outcomes that fall short of expectations.

Approach: Contributing to research on inclusive design principles for healthcare AI — examining how AI systems can be designed to serve diverse clinical users, with attention to transparency, adoption, and human-centred deployment.

Outcome: Ongoing research outputs and frameworks supporting responsible AI adoption in clinical settings.

→ Ongoing: contributing to human-centred frameworks for clinical AI adoption