1) Core Data Model

The app is built around a row-level “Minutes / Records” table. Each row is a discrete entry tied to a meeting via Meeting ID.

• ID (unique), Meeting ID, Date, Meeting Type
• Authority (name/code), Speaker, Role
• Cleaned Dialogue/Action, Topic/Keywords, Context/Notes
• Validation/Minutes Reference and Source/PDF/URL (when available)

2) Navigation Architecture

• Primary entry: searchable list/table of records
• Primary controls: search + filters (year, topic, speaker, authority)
• Detail view: expands context, validation, and sources
• Meeting-level grouping via Meeting ID supports deeper context review

3) Search & Tagging Strategy

Keyword normalization is used to track topics across time even as meeting language changes.

• Topic/Keywords (controlled categories for consistency)
• Additional Keywords (flexible tagging)
• Cleaned Dialogue/Action preserves meaning while improving readability

4) Validation & Transparency

• Source references are included when available
• Summaries are not official transcripts
• Dataset supports incremental improvement and correction over time

5) Permissions Philosophy

Designed for public research access with dataset integrity protected by restricting edits to authorized maintainers.

6) Mobile & Performance Priorities

• Mobile-first readability and scannable fields
• Filter-first workflow to reduce loaded records
• Context pushed into detail views to avoid list clutter

7) Expansion Path

The design supports future modules (budgets, capital projects, personnel timelines, authority dashboards) while keeping the same backbone: structured records + traceable sources + consistent tagging.