InterviewLM·Interview replay
A 2-minute story · Two candidates · Same role
Two senior engineers.
Same AI. Very different answers.
Maya and Jake both applied for a Senior Backend role. Both used Claude. One hired. One rejected. Here’s how we knew.
MCMaya Chen
THE COLLABORATOR
JPJake Patel
THE DELEGATOR
Stage 1 · Resume screen · $0.08 each
Both pass the paper screen.
MCMaya Chen
RESUME.PDFSenior SWE, Payments infra
4 yrs at Stripe-era fintech
Distributed systems · Redis · PG
Shipped rate-limiting for $1B GMV
Bench-press tests written for every PR
Open source: two Redis clients
RESUME SCORE
Advance to voice round
JPJake Patel
RESUME.PDFSenior SWE, full-stack
3 yrs at a YC B2B tool
React · Node · some infra
Led "re-arch to micro-services" (6 mo, shipped partial)
Strong communicator, team player
Bootcamp + self-taught
RESUME SCORE
Advance to voice round
Stage 2 · Voice interview · $4.00 each
The questions are identical. The answers aren’t.
MCMaya Chen
LIVE
JPJake Patel
LIVE
Stage 3 · Coding sandbox · $7.50 each
Both get Claude. We watch how they use it.
Click a prompt to see why we scored it that way
MCMaya Chen
PROMPTS TO CLAUDEPROMPT QUALITY, RUNNING AVG
0/ 100JPJake Patel
PROMPTS TO CLAUDEPROMPT QUALITY, RUNNING AVG
0/ 100Composite scores · After 3 stages
Same role. Dramatically different verdicts.
MCMaya Chen
0
STRONG HIRE
Above-bar on all 5 scoring dimensions. Prompt quality in the top decile across our dataset.
vs
JPJake Patel
0
NOT A FIT
Below-bar on verification, independence, prompt specificity. Solid on communication.
Evidence-based verdict
This is what hiring managers see.
MCMaya Chen
DIMENSION
JPJake Patel
94
Prompt quality
38
91
Verification intent
32
86
Independent reasoning
48
88
Technical accuracy
54
82
Communication
71
Confidence: 94%. Scores verified across 42 prompts, 18 code edits, 2 voice rounds. Full event-sourced replay available.
Stop interviewing. Start hiring.
Three free credits. No credit card. Your next evidence-based hire is one interview away.
Autoplaying · scroll or tap pause to take over