How CivicDeepDive Works

Step 1 — Smart parsing and entity extraction

Drop in a PDF of the minutes or agenda, or paste a public link to one. Our AI identifies who voted, what each resolution number is, and the exact public funds allocated.

• Resolution and agenda item numbers pulled from the original text.
• Dollar amounts extracted and totaled per meeting.
• Votes and passage status captured where the record states them.
• A link back to the official record so every line can be verified.

Step 2 — Prioritized categorization

The platform evaluates each item against civic patterns and tags it into Red-to-Blue warning tiers, so a critical policy shift never gets lost between routine business.

• Critical — big dollars, sole-source contracts, or direct public impact.
• Warning — rushed process, or money buried on the consent calendar.
• Watch — worth reading and tracking over time.
• Clear — routine administrative business.
• A transparency score per meeting, based on how much was bundled or left uncosted.

Step 3 — One-click mobilization

Every flagged item carries a direct path to action: match with neighbors tracking the same alert, draft letters to officials and board members, file a structured ombudsman complaint, or package a press kit for local reporters.

It compounds as the community grows

Each packet added makes the picture sharper. Agencies get compared to their peers, repeat patterns surface across jurisdictions, and spending anomalies stand out at a macro level that no single meeting ever reveals.