ADAPT
At Trinity College Dublin’s ADAPT Centre, I worked on collaborative projects with academic, clinical, and industry partners. My roles ranged from software engineering and architecture to team leadership.
The work covered digital health, natural language processing, recommender systems, computer vision, machine learning, and data visualisation. Here are some of the projects and public engagement activities I contributed to.
Highlighted projects
- Athena: research tools for studying information manipulation.
- PrecisionALS: secure data collection for rare disease research.
- Voicetune: speech synthesis evaluation tools.
Athena
Senior Research Engineer / Software Architect
Athena is a project focused on Foreign Information Manipulation and Interference (FIMI): intentional, coordinated behaviour by state or non-state actors aimed at distorting public perception, eroding trust, and undermining democratic institutions. ADAPT is creating a platform for researchers to work with FIMI, supporting dataset creation, annotation with NLP and AI tools, and dashboard-based exploration.
- I familiarised myself with a codebase developed before I joined the project.
- I set up automated builds using GitHub Actions.
- I deployed a second staging instance for testing.
Asthma Wearables
Senior Research Engineer / Software Architect
Asthma Wearables is developing a digital health platform that integrates wearable data, home spirometry, and patient-reported outcomes for continuous asthma monitoring and early detection of exacerbations.
- I reviewed patient data, identified data quality issues, and provided feedback to platform developers during data collection.
- I performed exploratory data analysis to support discussions with clinical stakeholders and guide research direction.
- I designed feature selection approaches, including KD-tree methods, to support modelling of asthma outcomes.
- I planned experimental frameworks for evaluating predictive signals from wearable and spirometry data.
ROBUST
Senior Research Engineer / Software Architect
ROBUST is a research project trialling electric shared mobility hubs across Ireland. It studies how e-cars, e-bikes, and e-cargo bikes can drive sustained changes in travel behaviour.
- I developed an iOS and Android app in React Native to show hub locations, collect survey responses, and gather usage data on shared mobility services.
PARADISE
Senior Research Engineer / Software Architect
PARADISE is a personalised medicine project focused on predicting relapse risk in ANCA-associated vasculitis. It combines clinical, biomarker, and patient-generated data to support AI-driven clinical decision tools, with an emphasis on semantic data integration and explainable AI.
- I carried out a security review of the existing development.
- I ported the web service to run locally as an Electron application so sensitive medical data could be processed on a clinician’s device.
PrecisionALS
Senior Software Engineer / Web Architect PrecisionALS is a collaborative research project between the School of Medicine and the School of Computer Science. I worked on the development and deployment of its data collection system, bringing initially outsourced Node.js back-end and Android app development in-house. I also worked on securing and deploying the back-end services on a university-hosted VM.
- I developed a test suite for an externally developed Node.js Express API.
- I made security improvements to pass a security audit.
- I deployed services on a college-administered VM, managing client certificates and two-way SSL to restrict access to a sensitive medical dataset.
- I managed a team of three junior developers and two interns across React, Node.js, and Android Jetpack Compose.
- I ported the data collection form rendering code from Android/Kotlin to React/JavaScript.
- I automated builds using GitHub Actions.
Machine Learning Dublin
Event Organiser
Machine Learning Dublin brought public talks on AI and machine learning to a local meetup audience.
- I created event templates and branding materials for the meetup.
- I coordinated speakers and catering for public events.
- I managed social media promotion to support event attendance.
Voicetune
Fullstack
Voicetune is an ADAPT spin-out funded by Enterprise Ireland. I supported its transition from academic research to a commercial spin-out. It focused on emotive text-to-speech for creative fields such as video game localisation and audiobooks.
- I designed a web app for researchers to evaluate speech synthesis models using different parameters.
- I implemented a FastAPI back-end service to serve models built with PyTorch.
- I implemented a Vue.js front end for evaluating trained models, including waveform and spectrogram visualisations.
- I implemented a proof-of-concept Vue.js script editing interface to demonstrate audio generation for game scenes and other multi-speaker dialogue.
HSE COVID Tracker App Review
Senior Research Engineer / Software Architect
ADAPT participated in an SFI review panel assessing the codebase for the COVID Tracker app, developed by NearForm.
- I reviewed the hosting and architecture of the app.
- I contributed to a report for the HSE on the app.
Time Series Anomaly Detection for DevOps
Machine Learning / Data Engineer Monitoring virtual machines, containers, databases, and other services is challenging: expected resource usage for one service might be a reason for concern for another. This project reproduced existing research and explored how domain adaptation could use data annotated for one system to detect possible faults in a target system without annotated data.
- I implemented a Python script to apply clustering, domain adaptation, and anomaly detection algorithms to a processed dataset and compare system performance.
- I implemented scripts to generate SLURM jobs for an HPC cluster and run experiment permutations to generate metrics.
Titles in the New Testament
Scrum Master
This project built an annotation tool to collect data stored in the New Testament Virtual Manuscript Room, a repository of manuscript scans built using Liferay. We extended existing widgets to support more complex annotations, including TEI transcripts.
- I supervised development of the project website and collection interface, mentoring a developer with less web development experience.
- I collaborated with a developer focused on data extraction to build an exploratory search interface.
Huawei: Adverts 2
Scrum Master This follow-on to Huawei: Adverts developed tools for natural-looking integration of 3D models into video. We developed a demo system for lens detection, surface recognition, perspective matching, and object replacement.
- I created a video dataset for object tracking and replacement.
- I oversaw delivery of the demo system, reusing the packaging and deployment process I had established in the previous project.
GAP: Gaming for Peace
Senior Software Engineer
GAP was an EU Horizon 2020 project developing a game to train peacekeepers, including army and police officers, in gender, cultural, and situational awareness. This training is typically delivered through in-person role-play.
- I developed a Unity plugin for event logging and syncing to a server.
- I developed a back end to store and process logged events.
- I developed a dashboard where users could view a report of their playthrough.
Huawei: Adverts
Scrum Master
This project developed a computer vision pipeline for billboard detection and advert replacement analysis in video. A C++ binary tracked, transformed, blended, and handled occlusion mapping to integrate a billboard or poster into video; an Express and Vue.js interface supported the workflow. The project received the 2018 Technology Ireland Industry Outstanding Academic Achievement Award.
- I coordinated image dataset creation and annotation.
- I oversaw pre-processing workflows for model training and evaluation, and development of API and CLI wrappers for recognition, localisation, and selection services.
- I produced installation and usage documentation, dockerised the deliverables, and assisted with deployment for industry partners.
Huawei: Semantic Understanding Video Summarisation
Scrum Master
This project focused on semantic video search and automatic summarisation for video collections. We developed annotation tools, knowledge-graph pipelines, and search tools to help users find objects, actors, actions, and moods within video archives.
- I coordinated technical delivery, including work allocation across the team.
- I contributed to team management responsibilities, including probation reviews and performance feedback.
Ryanair: Hotel Recommender System
Data / Software Engineer Ryanair was developing Ryanair Rooms, a hotel search and booking platform. With no hotel booking information available to rank listings, we developed a recommendation system using flight bookings and customer segmentation to bootstrap recommendations.
- I cleaned and normalised a provided data export, processing close to a terabyte of SQL exports.
- I created a demo web interface to mock a hotel search engine.
Avatar Trust
Research Engineer
Avatar Trust was an experiment at the Dublin Science Gallery. Participants interacted with a virtual avatar whose motion-captured face was projected onto a robot. The study explored whether a mismatch between the vocal track and the actor’s delivery affected trust in the avatar.
- I adapted an earlier tablet-based experiment to use a robot as the input device.
- I edited video from the earlier experiment for projection onto the robot while matching the motion in the original recordings.
AILO
Summer Internship Supervisor
The All Ireland Linguistics Olympiad (AILO) is an annual competition organised by the ADAPT Centre.
- I supervised summer interns who developed a web-based entry process for the preliminary-round exam.
DCU Fuse
Software Engineer
This project developed a website for a 24-hour public brainstorming session for DCU’s 2017 strategic plan. It received a Presidents’ Award from DCU.
- I wrote a plugin for Discourse, an open-source forum platform.
- I provided technical support during the event.
Brite:Bill
Software Developer
This project developed a machine learning proof of concept for personalised customer engagement workflows in telecom and utilities billing.
- I developed a synthetic data generator to create customer journey datasets for model development and validation.
- I developed an Apache Spark and Scala data pipeline codebase.
- I worked on-site with the industry partner, integrating with their sprint-based Agile workflow and delivery cadence.
RTÉ General Election 2016 Coverage
Software Engineer
This follow-on project used ADAPT NLP tools to analyse what people were discussing on Twitter during the general election.
- I updated the Twitter harvesting scripts.
- I designed and implemented a new back end and front end for interactive graphs.
- I implemented graphs that were used in RTÉ News.
Intel: Localisation Bus
Software Engineer
The Enterprise Localisation Bus project built an integration architecture for multilingual content workflows using OASIS XLIFF and CMIS standards. The platform combined CMIS, MuleSoft, and Vue.js to support translation and content movement across systems.
- I developed MuleSoft pipelines to call machine translation services.
- I developed a WordPress plugin to expose content through CMIS APIs and trigger localisation workflows.
- I developed a dashboard to show data flow through the localisation process.