The TechIreland National AI Challenge 2026 ran under the theme “AI & Agents: Strategy, Processes and Roles”. Over two weeks, developers, entrepreneurs, researchers, students, and professionals worked together to solve real-world industry challenges using AI and agentic AI technologies across sectors from manufacturing to healthcare to public services.
Tus-Aite is our entry on the healthcare track. It targets the problem of outpatient waiting list prioritisation in the Irish health system: an AI agent that sits between a national waiting list and the clinician who signs each referral — ranking patients within their urgency category using objective clinical signals and ward-level capacity, and explaining every position in plain language.
The waiting list is a synthetic set of Irish outpatient referrals, already sorted into Urgent, Semi-Urgent, and Routine per the NTPF national structure, with clinical detail and capacity invented for the challenge. The agent runs an Urgency tool (NEWS2 from latest vital signs), a Capacity tool (ward and clinic pressure), and a Coordinator tool that combines both scores, re-ranks within each category, and checks the five NTPF rules. The resulting Ranked list is saved with its evidence in the knowledge graph, and a Rationale tool explains each position from cited evidence. The clinician sees each position and its plain-language why, and can accept or override.
Code: View on GitHub.