Sovereign AI use case · Government and Public Sector

AI for public consultation analysis in government

Coding is often split across several people, which makes themes inconsistent. Long technical submissions from organizations and short individual comments are hard to weigh fairly. Summaries can be challenged when stakeholders cannot see how their input was considered. The cost is delay, contractor spend and a weaker decision record.

Inaaya use-case library · Last reviewed 19 November 2026

In short

Public consultation analysis is the reading and coding of submissions received in a consultation, grouping them by theme, counting positions and selecting representative quotes. Every finding stays linked to the submissions that support it. Officials decide how input shapes policy and what is published.

Who does this today.

Consultation and engagement teams, policy analysts and external contractors do this work today. It is hard because a consultation can draw hundreds or thousands of submissions, from short emails to long technical briefs, and coding them consistently by hand is slow and hard to verify.

What Inaaya can do, step by step.

  1. Reads every submission in the formats received, including written messages, forms and attached briefs.
  2. Separates submissions from individuals, organizations and governments as the department defines. Submissions from Indigenous governments and organizations are flagged to the team responsible for Indigenous consultation, which is handled separately from public comment analysis.
  3. Groups comments by theme and by the consultation questions asked.
  4. Counts support, opposition and conditional positions on each proposal.
  5. Pulls representative quotes with a link to the source submission.
  6. Flags duplicate or campaign submissions for the team to decide how to treat.
  7. Drafts a What We Heard summary for officials to edit.
You provide
  • Consultation submissions
  • The consultation questions and proposals
  • The department's coding and reporting approach
You get
  • A coded submission register
  • Theme and position counts
  • A draft What We Heard report with quotes and sources
Your people decide
  • Consultation leads review the coding framework
  • Officials verify theme summaries
  • Publication approved by the department
Data and controls
  • Can run with Canadian data residency on a model and host the department approves
  • Every action recorded with who, what and when, for audit and access requests
  • Personal information in submissions handled under applicable privacy rules
  • Quotes anonymized where the department requires
How to measure it
  • Time from consultation close to What We Heard report
  • Share of submissions coded and traceable
  • Stakeholder challenges to the summary
The aim

Designed so that the department has a transparent, verifiable summary of what people said, ready for the decision record.

Questions people ask.

Can AI analyze public consultation submissions?

Inaaya can read every submission, group themes and count positions, keeping a link to each source so the summary can be checked.

How are campaign submissions handled?

They are flagged as duplicates or campaigns, and the team decides how to count and report them.

Does AI decide which input to act on?

No. It summarizes what was said. Officials decide how the input informs policy.

Related use cases.

Apply to be a founding partner