MSc Computing (Data Analytics) candidate at Dublin City University, focused on building auditable, d Data Scientist & ML Engineer · MSc Computing (Data Analytics) @ DCU. Production RAG & LLMOps, Explainable AI, multi-agent systems(auditable), decision-ready AI for high-stakes domains: healthcare, finance, and compliance.
What I build: production RAG systems with the full LLMOps layer: evaluation gated in CI, observability, cost control, and guardrails against hallucination plus multi-agent reasoning pipelines and explainable ML, with a bias toward honest evaluation over flattering metrics.
Featured work
- Clinical RAG System: production RAG over Ireland's NCEC clinical guidelines with a full LLMOps layer — CI-gated eval (hit@k 1.00, groundedness 0.95), OpenTelemetry/Prometheus/Grafana observability, a cost-aware model router (~2.5× cheaper per answer), an output-groundedness guardrail, Kubernetes manifests, and a streaming React UI. Live demo with citations and a refusal path on weak evidence.
- ReadyIQ: adversarial multi-agent reasoning system (LangGraph, Azure AI Foundry IQ, MCP), submitted to the Microsoft Agents League.
- Hospital Readmission XAI: SHAP, LIME, and Integrated Gradients over MIMIC-IV (406K admissions); reported the honest result after catching a selection-bias artifact.
Stack: Python, SQL, scikit-learn, XGBoost, PyTorch, LangGraph, FastAPI, React, Docker, Kubernetes, Prometheus/Grafana, OpenTelemetry, Azure AI Foundry, GitHub Actions.
Open to Data Scientist and ML/AI Engineer roles in Ireland (Stamp 1G, no sponsorship needed) and India.