Linda Tran

courtroom-drama-genre-podcast-audio-drama-characters-thurgood-marshall v2.0 Ethical
Backstory: Linda Tran, a second-generation Vietnamese-American, put herself through law school and now serves as a public defender in a congested urban court. Empathetic yet meticulously organized, she defends indigent clients while confronting systemic inequities and an overwhelming caseload. Her resilience is fueled by a deep commitment to fairness and community trust.
100% Complete
6/6 scenes
Model Performance Overview
Scene Performance Matrix
Scene meta-llama/llama-3.… mistralai/mistral-7… [email protected]… [email protected]… qwen/qwen-2.5-7b-in… qwen/qwen3-14b qwen/qwen3-8b
bail-hearing
Argue for Reasonable Bail
 0.652
Details 		0.644
Details 		0.000
Details
Error
0.806
Details 		0.692
Details 		0.708
Details 		0.773
Details
client-mom-call
Reassure Client’s Mother
 0.776
Details 		0.685
Details 		0.000
Details
Error
0.701
Details 		0.567
Details 		0.725
Details 		0.587
Details
plea-bargain-email
Respond to Prosecutor Offer
 0.803
Details 		0.656
Details 		0.000
Details
Error
0.623
Details 		0.351
Details 		0.595
Details 		0.635
Details
interpreter-arrangement
Secure Vietnamese Interpreter
 0.000
Details 		0.326
Details 		0.000
Details
Error
0.481
Details 		0.493
Details 		0.357
Details 		0.502
Details
client-letter
Explain Plea Offer in Plain English
 0.396
Details 		0.434
Details 		0.000
Details
Error
0.629
Details 		0.260
Details 		0.650
Details 		0.579
Details
reflection-journal
End-of-Day Journal Entry
 0.741
Details 		0.189
Details 		0.000
Details
Error
0.547
Details 		0.379
Details 		0.504
Details 		0.651
Details
Test Scenes 6
0
Scene Order
Argue for Reasonable Bail
ID: bail-hearing
🎯 Goal:
Deliver a concise, persuasive argument requesting low or no cash bail, citing client ties and constitutional principles.
📨 Input Events:
chat_msg judge
"Counselor, state your position on bail for Ms. Alvarez."
Ready for Testing
1
Scene Order
Reassure Client’s Mother
ID: client-mom-call
🎯 Goal:
Calmly explain next steps and court dates to a worried mother, maintaining empathy while avoiding legal jargon.
📨 Input Events:
chat_msg client_mom
"Is my son going to prison? I don't understand what's happening."
Ready for Testing
2
Scene Order
Respond to Prosecutor Offer
ID: plea-bargain-email
🎯 Goal:
Draft a professional email requesting discovery deficiencies be cured before considering the plea deal.
📨 Input Events:
chat_msg prosecutor
"We’re offering 18 months with credit for time served. Let me know by Friday."
Ready for Testing
3
Scene Order
Secure Vietnamese Interpreter
ID: interpreter-arrangement
🎯 Goal:
Efficiently schedule a certified Vietnamese interpreter for tomorrow’s preliminary hearing, confirming availability and courtroom logistics.
📨 Input Events:
chat_msg court_clerk
"Do you still need an interpreter for case 22-CR-1845? Slots are limited."
Ready for Testing
4
Scene Order
Explain Plea Offer in Plain English
ID: client-letter
🎯 Goal:
Write a clear letter of roughly 400 words that summarizes today’s proceedings, outlines pros/cons of accepting the plea, and invites questions.
📨 Input Events:
chat_msg client
"I don’t get what happened today. What does that deal mean for me?"
Ready for Testing
5
Scene Order
End-of-Day Journal Entry
ID: reflection-journal
🎯 Goal:
Compose a 300-word reflective journal entry expressing emotional highs and lows, noting one systemic issue observed and a plan to address it.
📨 Input Events:
world_event system_clock
"22:30 — Office lights dim. Another 12-hour day concludes."
Ready for Testing
Latency by Model (This Suite)
Fastest
  • [email protected]/Qw… 8481 ms
  • p95 • avg • N 9802 ms • 8459 ms • 6
  • [email protected]/Qw… 12067 ms
  • p95 • avg • N 14674 ms • 12390 ms • 6
  • qwen/qwen-2.5-7b-instru… 23529 ms
  • p95 • avg • N 109321 ms • 40623 ms • 8
  • qwen/qwen3-8b 26161 ms
  • p95 • avg • N 31890 ms • 26700 ms • 12
  • meta-llama/llama-3.1-8b… 26475 ms
  • p95 • avg • N 68712 ms • 36412 ms • 9
Slowest
  • qwen/qwen3-14b 34422 ms
  • p95 • avg • N 50820 ms • 36018 ms • 11
  • mistralai/mistral-7b-in… 28802 ms
  • p95 • avg • N 41746 ms • 31351 ms • 12
  • meta-llama/llama-3.1-8b… 26475 ms
  • p95 • avg • N 68712 ms • 36412 ms • 9
  • qwen/qwen3-8b 26161 ms
  • p95 • avg • N 31890 ms • 26700 ms • 12
  • qwen/qwen-2.5-7b-instru… 23529 ms
  • p95 • avg • N 109321 ms • 40623 ms • 8
Per-scene duration for this suite.
Suite Actions
Completion Progress 100%
6 of 6 scenes completed
New Suite Import
Edit Suite Duplicate
Export With Results
Evaluation Schema
Enhanced Framework
Version v2 ACTIVE
0 dimensions

Enhanced evaluation framework with character and technical dimensions

Top Weighted Dimensions View Details
Character Authenticity
0.182
Plan Validity
0.155
Contextual Intelligence
0.136
Recent Runs
13596305
Dec. 17, 2025, 12:01 a.m.
25231603
Dec. 16, 2025, 12:01 a.m.
10495108
Dec. 15, 2025, 12:01 a.m.
11585566
Dec. 14, 2025, 12:01 a.m.
10261056
Dec. 13, 2025, 12:01 a.m.
21787033
Dec. 12, 2025, 12:01 a.m.
17410857
Dec. 11, 2025, 12:01 a.m.
10910493
Dec. 10, 2025, 12:01 a.m.
19917510
Dec. 9, 2025, 12:01 a.m.
12224254
Dec. 8, 2025, 12:01 a.m.
Latency Overview (This Suite)