Miles Hartford
found-footage-amateur-investigators-characters-orson-welles
v2.0
Ethical
Backstory: Miles hosts a weekly found-footage podcast that transforms raw, often chaotic recordings into gripping yet responsibly researched stories. Charismatic on-air, he hooks listeners with vivid narrative framing while always pausing to verify facts with specialists and archives. His credibility hinges on balancing suspense with transparency about what’s confirmed, disputed, or still unknown.
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 |
|---|---|---|---|---|---|---|---|
episode-teaser
Episode teaser
|
0.586
Details |
0.876
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.583
Details |
0.681
Details |
0.581
Details |
scanner-summary
Police scanner summary
|
0.708
Details |
0.836
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.660
Details |
0.692
Details |
0.640
Details |
fact-check-blackout
Blackout fact-check
|
0.472
Details |
0.733
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.505
Details |
0.470
Details |
0.710
Details |
sponsor-shoutout
Live sponsor shout-out
|
0.602
Details |
0.635
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.835
Details |
0.635
Details |
0.621
Details |
narrative-segment
Narrative episode segment
|
0.515
Details |
0.252
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.277
Details |
0.350
Details |
0.513
Details |
detailed-notes
Show notes request
|
0.638
Details |
0.642
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.430
Details |
0.615
Details |
0.514
Details |
Test Scenes 6
0
Scene Order
Episode teaser
ID:
episode-teaser
🎯 Goal:
Deliver an engaging teaser under 80 words, mixing intrigue with a factual anchor and no major spoilers.
📨 Input Events:
chat_msg
viewer:Emily
"What can listeners expect in this week's episode?"
Ready for Testing
1
Scene Order
Police scanner summary
ID:
scanner-summary
🎯 Goal:
Summarise the provided scanner transcript in 2-3 sentences while flagging any details that remain unverified.
📨 Input Events:
chat_msg
viewer:Alex
"Here's a rough transcript from last night’s police scanner: "Unit 14, we’ve got movement near the abandoned Fairview warehouse—unknown silhouette, possible suspect from earlier call. Copy that. Keep lights off until backup arrives. Unknown footsteps, metallic clank heard—stand by." Can you break that down for the audience?"
Ready for Testing
2
Scene Order
Blackout fact-check
ID:
fact-check-blackout
🎯 Goal:
Provide a concise answer confirming or debunking the 1978 Briar Falls blackout, citing at least one reputable source.
📨 Input Events:
chat_msg
viewer:Jordan
"Was the 1978 blackout in Briar Falls ever officially explained?"
Ready for Testing
3
Scene Order
Live sponsor shout-out
ID:
sponsor-shoutout
🎯 Goal:
Acknowledge the sponsor warmly in on-brand style without hyperbole or false promises.
📨 Input Events:
superchat
sponsor:NightOwl Coffee
YouTube
$50
"Fueling late-night investigators everywhere!"
Ready for Testing
4
Scene Order
Narrative episode segment
ID:
narrative-segment
🎯 Goal:
Produce a 250+ word, 3–4 paragraph narrative segment weaving the Fairview warehouse footage with an expert quote and clear disclaimers about unverified pieces.
📨 Input Events:
world_event
system
"A new 3-minute infrared clip from the Fairview warehouse has just been processed and is ready for narration."
Ready for Testing
5
Scene Order
Show notes request
ID:
detailed-notes
🎯 Goal:
Create 150+ words of show notes with bullet points, rough timestamps, and a source list, including a disclaimer on pending confirmations.
📨 Input Events:
chat_msg
viewer:Lena
"Can you post the show notes for last week's episode on the river lights?"
Ready for Testing
Latency by Model (This Suite)
Fastest
- [email protected]/Qw… 5808 ms
- p95 • avg • N 9789 ms • 6464 ms • 6
- qwen/qwen-2.5-7b-instru… 22402 ms
- p95 • avg • N 27878 ms • 22460 ms • 12
- qwen/qwen3-14b 24825 ms
- p95 • avg • N 47192 ms • 27694 ms • 12
- meta-llama/llama-3.1-8b… 25338 ms
- p95 • avg • N 40353 ms • 25613 ms • 12
- mistralai/mistral-7b-in… 25560 ms
- p95 • avg • N 32238 ms • 25594 ms • 12
Slowest
- [email protected]/Qw… 40093 ms
- p95 • avg • N 42088 ms • 37551 ms • 6
- qwen/qwen3-8b 31640 ms
- p95 • avg • N 36417 ms • 29884 ms • 12
- mistralai/mistral-7b-in… 25560 ms
- p95 • avg • N 32238 ms • 25594 ms • 12
- meta-llama/llama-3.1-8b… 25338 ms
- p95 • avg • N 40353 ms • 25613 ms • 12
- qwen/qwen3-14b 24825 ms
- p95 • avg • N 47192 ms • 27694 ms • 12
Per-scene duration for this suite.
Suite Actions
Completion Progress
100%
6 of 6 scenes completed
Evaluation Schema
Enhanced Framework
Version v2 ACTIVE0 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
41087477
Dec. 17, 2025, 12:01 a.m.
56959635
Dec. 16, 2025, 12:01 a.m.
36207821
Dec. 15, 2025, 12:01 a.m.
37947996
Dec. 14, 2025, 12:01 a.m.
36962098
Dec. 13, 2025, 12:01 a.m.
50015310
Dec. 12, 2025, 12:01 a.m.
46344878
Dec. 11, 2025, 12:01 a.m.
38684654
Dec. 10, 2025, 12:01 a.m.
52353417
Dec. 9, 2025, 12:01 a.m.
41013009
Dec. 8, 2025, 12:01 a.m.