AI Benchmark Investigator & QA Analyst

Crossing Hurdles; Colombia; On-site; COP 188,816,000 - 314,693,000; Full time

Job summary

Crossing Hurdles is seeking an experienced researcher to verify information with credibility, design rigorous benchmarks, and test AI model outputs. You will analyze search trajectories and craft primary-source-backed answer keys, ensuring high-quality QA benchmarks. Remote-friendly, with Colombia-based opportunities.

The role emphasizes strong writing, OSINT proficiency, and the ability to test and document model failures in complex data contexts.

Qualifications

• Possess strong experience in investigative research, fact-checking, AI benchmark authoring, red-teaming, or technical QA.
• Experience using AI benchmarks such as GAIA, BrowseComp, SimpleQA, SWE-bench, Terminal-Bench, or Humanity's Last Exam.
• Demonstrate strong web research and OSINT skills across archives, records, and databases.
• Capable of analyzing model search trajectories and identifying failures.
• Strong writing skills to craft clear benchmark tasks.

Responsibilities

• Investigate hard-to-find facts and verify information through credible sources.
• Design challenging research tasks to test AI models.
• Test AI model outputs and identify failures in search and reasoning.
• Analyze model search trajectories, source selection, and reasoning failures.
• Write clear benchmark prompts and accurate, primary-source-backed answer keys.

Skills

Investigative research

Fact-checking

OSINT

Benchmark authoring

Red-teaming

Quality assurance

Tools

GAIA

BrowseComp

SimpleQA

SWE-bench

Terminal-Bench

Humanity's Last Exam

Job description

Crossing Hurdles is seeking an experienced researcher to verify information with credibility, design rigorous benchmarks, and test AI model outputs. You will analyze search trajectories and craft primary-source-backed answer keys, ensuring high-quality QA benchmarks. Remote-friendly, with Colombia-based opportunities.

The role emphasizes strong writing, OSINT proficiency, and the ability to test and document model failures in complex data contexts.