← back to feed
Autopsy complete · re-runs every 24h
openai

Data Scientist, Cybersecurity @ OpenAI

remote·seniority unknown·today
88/100
verified real
"Massive pay to figure out if our autonomous agents are going rogue or just being annoying."
— The Judge · model: gemini-3.5-flash · confidence: high
01

The five agents · full report

// parallel · uncensored · re-runnable
The Appraiser
comp inference
$263k–$515kUSD / yr
confidencehigh
disclosed by employerNO — penalty applied
posted band$263k–$515k
The Skeptic
red & green flags
  • At-will clause notes they can modify base pay components at any time
  • Highly ambiguous 'agentic security' frontier with limited industry precedents
  • Spectacular and fully transparent base salary range
  • Clear, high-ownership scope over a novel cybersecurity domain
  • Generous benefits including up to 24 weeks of parental leave
The Decoder
stack + team maturity
PythonSQLLLMCausal Inference
data teamStage 3 · Platform
company stage
size
The Oracle
interview prediction
rounds5
formatTechnical coding, security analytics case study, and system design
unpaid take-homenone
02

What you can do about it

// apply · flag · negotiate · ignore
▼ NO APPLY URL
Employer didn't provide one. Hide-and-seek loses you 10 BS pts.
🔒 NEGOTIATION COACH · PRO
Live script · counter-offer drafts · expected ranges.
04

Reply Guy

// agent-drafted cover letter · cursedly on-brand
Vibe:3 of 3 free left
05

Ghost-job market

// stake XP · settled in 60 days

Bet ghost? This posting still open in 60 days with no signs of a hire — you win the pool. Bet real? Win if it closes with someone hired.

▼ ghost 0 XP · 0 bettors▲ real 0 XP · 0 bettors
Stake:
How this autopsy was generated

5 parallel AI agents (Gemini 3 Flash) extract structured signals from the JD, cross-reference with our company graph, infer market comp, predict the interview loop, and synthesize a final BS Score. Re-run every 24 hours. Employers can request a re-autopsy after editing the JD — no charge, no special access.