# AI Brady Discovery for Prosecutors

*/Opportunities/AI_Brady_Discovery_for_Prosecutors*

## Opportunity Market Sizing

_Illustrative — target and order-of-magnitude estimate figures, not an achieved track record (this Thing is concept-stage)._

**S A M**: ~$200-300M targeting mid-to-large metropolitan District Attorney and US Attorney offices
**S O M**: ~$15-30M
**T A M**: ~3,000 US state, local, and federal prosecutor offices × ~$150k/yr average discovery software spend ≈ $450M
**Growth Rate**: ~12-18%/yr, driven by the exponential explosion of digital case evidence compounding Brady compliance risks
**Paid Comparable Spend**: Typically ~$100k-500k/yr per office spent on paralegal overtime, contract document reviewers, and legacy eDiscovery platforms to manually screen bodycam footage and phone dumps

## Opportunity Incumbents

- [Relativity eDiscovery](/Products/Relativity_eDiscovery) — Tool
- [Everlaw Platform](/Products/Everlaw_Platform) — Tool
- [Prosecutor By Karpel](/Products/Prosecutor_By_Karpel) — Tool
- [Tyler Odyssey](/Products/Tyler_Odyssey) — Tool
- [Manual Paralegal Review](/Products/Manual_Paralegal_Review) — Service
- [Excel Disclosure Logs](/Products/Excel_Disclosure_Logs) — Spreadsheet

## Opportunity Win Conditions

**Kill Thresholds**:
- False positive rate > 30% after 30 days of model tuning
- Pilot-to-paid conversion < 20% after 90 days
- Time-to-first-value > 72 hours due to ingestion bottlenecks
- CJIS compliance certification blocked or delayed beyond 60 days
**Leading Metrics**:
- Time from evidence upload to generated disclosure log
- False positive rate on flagged exculpatory events
- Weekly active days per trial attorney
- Percentage of AI flags manually dismissed by paralegals
**What Proves Right**: Prosecutors upload bodycam footage and phone extraction files to the system and review automated exculpatory evidence flags within four hours. Mid-sized DA offices convert 30-day pilots into paid annual contracts at the $150,000 price point. Weekly active usage among trial attorneys exceeds 75% as they export the generated disclosure logs directly into case files.
**What Proves Wrong**: Trial teams reject the automated flags because missing a single piece of exculpatory evidence carries career-ending malpractice risks. The engine generates a false positive rate above 40%, forcing paralegals to spend more time clearing bad flags than they would manually reviewing the raw files. Municipal IT departments block deployment due to hard blockers on cloud-based CJIS compliance.

## Neighborhood

### Entrant startups

- [Compass](/Startups/Compass) — is entrant in · Startups

### Incumbent in

- [Everlaw Cloud Platform](/Products/Everlaw_Cloud_Platform) — incumbent in · Products
- [Excel Disclosure Logs](/Products/Excel_Disclosure_Logs) — incumbent in · Products
- [Manual Paralegal Review](/Products/Manual_Paralegal_Review) — incumbent in · Products
- [Prosecutor By Karpel](/Products/Prosecutor_By_Karpel) — incumbent in · Products
- [Relativity eDiscovery](/Products/Relativity_eDiscovery) — incumbent in · Products
- [Tyler Odyssey](/Products/Tyler_Odyssey) — incumbent in · Products