Fraud AI Engineer & Frontier Code Evaluator

Obsidian; Boston (MA); On-site; USD 454,608 - 647,472; Part time

Job summary

Obsidian is seeking professionals for a unique role focused on improving frontier AI coding models in fraud and risk engineering. As a contributor, you will evaluate model outputs and contribute to the development of fraud detection systems and risk scoring models.

Applicants should have a minimum of 2 years of professional experience in relevant fields and familiarity with AI coding agents. Compensation is $400 per accepted task, with each task taking about 2-3 hours.

Qualifications

• 2+ years of professional experience in fraud engineering or related fields.
• Experience with fraud detection systems and risk scoring models.
• Familiarity with AI coding agents.

Responsibilities

• Use AI coding agents to evaluate complex fraud and risk engineering tasks.
• Identify bugs and failure modes in model implementations.
• Compare outputs from different AI models.

Skills

Fraud engineering

Risk engineering

Trust & safety

AI coding agents

Tools

Cursor

Claude Code

Codex

Job description

Obsidian is seeking professionals for a unique role focused on improving frontier AI coding models in fraud and risk engineering. As a contributor, you will evaluate model outputs and contribute to the development of fraud detection systems and risk scoring models.

Applicants should have a minimum of 2 years of professional experience in relevant fields and familiarity with AI coding agents. Compensation is $400 per accepted task, with each task taking about 2-3 hours.