I build AI agents, RAG systems, and production ML pipelines — from raw data to deployed applications. Most recently at Cybereason, building AI for enterprise security analytics. MSc Data Science & Analytics, UCC.
I'm a Data Scientist and AI Engineer who most recently spent 2.5 years at Cybereason building production ML and AI systems for enterprise security analytics. My work sits at the intersection of applied AI, data engineering, and backend development.
My background spans LLM agents, RAG pipelines, anomaly detection, and scalable data infrastructure — all built for high-stakes environments where reliability and observability matter.
The cybersecurity context isn't incidental — it means I've built AI systems that operate on messy, high-volume, adversarial data. That sharpens everything.
I'm comfortable owning the full stack: from raw data to deployed model to API to dashboard. If there's a gap in a system, I'll fill it.
Core Stack
Domain Coverage
2.5 years progressing from Graduate Analyst to Data Scientist at a global cybersecurity company.
Upload PDFs, ask questions in natural language, get accurate answers with page-level source citations. Supports 6 AI providers (Anthropic, OpenAI, Google, Mistral, Cohere, Groq) and 15+ models. Built with a FastAPI backend doing real RAG — chunking, embedding, and ChromaDB vector search — with a clean zero-dependency frontend. Users bring their own API key, keeping operating costs at zero.
AI-powered case management platform for personal injury law firms. Features real PDF/DOCX text extraction in-browser, Claude-powered document analysis, a context-aware AI assistant with full case awareness, deadline tracking by urgency, and a document querying mode for deep-diving into specific files.
MSc thesis project benchmarking five face detection models — DLIB, MTCNN, FaceNet, RetinaFace, and OpenCV Caffe — for gender prediction on YouTube video frames. Built a full pipeline to download videos, extract frames, detect faces per frame, and classify gender using DeepFace. Evaluated accuracy, inference speed, and robustness across models on real-world ad footage.
I'm interested in roles building AI agents, LLM applications, or production ML systems — especially where the work is technically deep and the stakes are real.
Also open to conversations about applied GenAI, data engineering, or anything at the intersection of security and AI.
RESPONSE TIME — usually within 24 hours
PREFERRED — Email or LinkedIn DM
LOCATION — Ireland. Open to remote roles and relocation for the right opportunity.