Skip to main content
lab-notes.aiPublic-service AI playbooks
Opportunity 06/17HealthHealth and care

Health Service Demand and Operations

Explore which published figures could help a hospital planning team understand demand, beds, and discharge, and where the evidence remains incomplete.

Published sources
2
Starter data
20 synthetic records
Domain build partner
build-health-operations
Last reviewed

From the draft strategy

Opportunity

The draft strategy names AI support for health operational processes as a potential public-service application: help with discharge coordination, bed demand, and how limited capacity is shared out, without saying what the system would be aiming at or what it must never do.

Where the draft says it

Table 2 — potential public-service applications

Read the draft strategy source (opens in a new tab)

Source register

Research already done

Real published source

We reviewed these published sources so you do not have to start from zero. Each source shows what it covers, how it can be accessed, and why it matters.

2 sources

  1. Department of Health

    Open

    Hospital waiting times statistics

    What it covers
    Quarterly outpatient, inpatient and day case, diagnostic, and cancer waiting figures for Northern Ireland, broken down by health and social care trust and by how long people have waited.
    Why it matters here
    It is the published shape this playbook's synthetic records imitate: a specialty, a period, a length-of-wait band, and a count of people, and nothing about any one person.

    Open the source (opens in a new tab)

  2. Northern Ireland Statistics and Research Agency

    Open

    Health and social care statistics

    What it covers
    The wider collection of Northern Ireland health statistics, including hospital activity and waiting lists, primary care, social care, workforce, and health inequalities.
    Why it matters here
    It shows how much of the picture a planning team would need sits in separate publications, which is most of the real work in this example.

    Open the source (opens in a new tab)

Starter data

Starter dataset

Synthetic working data

Use 20 synthetic waiting records to explore a hospital planning question without holding health and social care trust data.

AI-assisted research helped identify and interpret the published sources. We then created a small, non-sensitive synthetic dataset shaped by the information those sources expose. It is for exploration—not evidence, training, or operational decisions.

Records
20
Fields
5

Field preview

table view
  • id

    SYN-HO-01 · SYN-HO-02 · SYN-HO-03

  • specialty

    General surgery · Trauma and orthopaedics · Ear, nose and throat

  • quarter

    2025-Q3 · 2025-Q4 · 2026-Q1

  • waitingBand

    0-6 weeks · over 52 weeks · 14-26 weeks

  • patientsWaitingBand

    2500-5000 · 10000-20000 · 5000-10000

How it was prepared

AI authored fictional specialty, quarter, wait-band, and count-band values in the shape of published statistics.

Limitations

  • The numbers are invented bands, not rounded real figures, so nothing here should be quoted as how long anyone in Northern Ireland is waiting.
  • Real planning needs arrivals, beds, staffing, and discharge delays together, and this file has only one of those.
  • A tidy quarterly table hides the day-to-day movement and the individual circumstances that actually decide when someone leaves hospital.

Clone and build

Domain build partner

Build with a domain-aware coding agent

Clone the repository, then ask your coding agent to load the skill that ships with it. The skill brings the opportunity, sources, starter data, known unknowns and constraints into the conversation before anything is proposed. It follows the open Agent Skills standard, so it works in Claude Code, Codex and any other agent that reads skills.

01 / Clone

git clone https://github.com/Hypership-Software/lab-notes.ai.git

02 / Invoke

Claude Code
/build-health-operations
Codex
$build-health-operations

Conditions, not a checklist

Before you build

These are not footnotes. They are conditions that any responsible prototype would need to address.

  • Operational records are sensitive

    Operational records describe people who are ill, so the real version of this work involves sensitive information even when the output looks like a chart.

  • Single targets can distort care

    Aiming at one number, such as average flow, can quietly make things worse for people whose needs are complicated or unusual.

  • Discharge decisions remain clinical

    Knowing how many people are likely to arrive does not tell anyone whether a particular person is ready to go home, and that decision has to stay with clinicians.