Skip to main content
lab-notes.aiPublic-service AI playbooks
Opportunity 12/17Cross-governmentCitizen services and government

Policy Evidence and Consultation Analysis

Explore how consultation responses might be grouped into themes for further investigation while keeping the supporting passages visible.

Inspect starter dataSet up the build partner
Published sources
2
Starter data
20 synthetic records
Domain build partner
build-policy-evidence
Last reviewed

From the draft strategy

Opportunity

The draft strategy names AI-assisted analysis of public consultation responses as a potential public-service application: helping policy teams see the themes in large volumes of free-text replies.

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. The Executive Office

    Open

    Northern Ireland Artificial Intelligence Strategy consultation

    What it covers
    The draft strategy text and the consultation it is open for.
    Why it matters here
    It is the document whose example projects these playbooks explore.

    Open the source (opens in a new tab)

  2. Department for the Economy

    Open

    Draft Circular Economy Strategy — public consultation response report

    What it covers
    How an NI department actually analysed and reported a consultation's responses.
    Why it matters here
    Its headings, stages, and vocabulary shaped the synthetic dataset's structure; no respondent text was copied. Our reading of it is recorded in consultation-analysis-structure.md beside this file.

    Open the source (opens in a new tab)

Starter data

Starter dataset

Synthetic working data

Use 20 synthetic consultation responses to explore theme grouping without holding a consultation mailbox.

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
4

Field preview

records view
  • id

    SYN-0001 · SYN-0002 · SYN-0003

  • theme

    access-to-services · workforce-capability · data-governance

  • stance

    critical · supportive · uncertain

  • text

    We work with people who have no home internet and no smartphone. The proposal assumes a digital route is a choice. For the people we support it is not, and the paper alternative has been quietly withdrawn from three of the offices they used to be able to walk to. · Anything that stops people telling the same story four times to four different teams is welcome. Please make sure the single record can be corrected by the person it describes, and not only by staff. · It is hard to comment without knowing what happens at the point of refusal. If a decision is reached faster but the review afterwards still takes eleven weeks, we are not sure the change helps the people who most need it.

Inspect all 20 recordsView repository file (opens in a new tab)Download JSON (opens in a new tab)

How it was prepared

AI authored fictional responses shaped by the structure and vocabulary of a published consultation response report.

Limitations

  • The dataset is far smaller and tidier than a real consultation mailbox.
  • The six themes and four stances are this project's own choices, verified in no official source.

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-policy-evidence
Codex
$build-policy-evidence

Conditions, not a checklist

Before you build

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

  • Keywords do not establish meaning

    A matched keyword shows a response used a word, not what the respondent meant by it; a real analysis needs human reading.

  • Synthetic analysis proves no operational outcome

    Nothing on this page is evidence that an AI system would analyse a real consultation accurately, fairly, or lawfully.