AI Courtroom Simulator
A native iOS simulation that generates structured criminal cases and lets a user move through preparation, examination, argument, deliberation, and verdict workflows.
01 / The problem
What needs to change
Legal reasoning and courtroom procedure are difficult to explore interactively without a complete case, varied participants, opposing arguments, and a persistent trial record.
02 / The approach
A focused path through the complexity
The app uses structured model output to create varied case facts and characters, then coordinates user and AI roles across a staged trial with saved transcripts, notes, and jury deliberation.
03 / Key capabilities
What the system is designed to do
- Structured case and participant generation
- Choice of legal role and model
- Pre-trial character conversations and case preparation
- Opening, examination, cross-examination, and closing stages
- AI jury deliberation, verdict, notebook, and transcript
04 / Architecture
Technology and system shape
- SwiftUI application with Core Data persistence
- Function-shaped model output for reliable case creation
- Keychain-backed credential handling
- State-driven trial stages and reusable role-specific views
- Network monitoring, transcripts, and local case history
05 / Results
What the repository demonstrates
- The prototype implements the full simulation arc from generated case through deliberation and verdict.
- Case characters, trial events, user notes, role choices, and progress persist across the native app experience.
06 / Lessons
What the work teaches
- Structured generation is essential when every later interaction depends on a coherent cast, evidence set, and ground truth.
- A simulation feels more credible when AI behavior is constrained by procedural stages and persistent context.
07 / Screenshots
Approved visuals
No public screenshots yet
Add only approved screenshots to the project record.