Member of Technical Staff

Remote in United States, Canada or Latin America • $200K-320K + Equity

Member of Technical Staff - AI Benchmark Curation & Validation

Role Summary

Own the quality of OB-1's benchmark suite. Execute tasks with the AI agent, analyze results, identify broken or gamed benchmarks, and curate hundreds of tasks for production. You need deep technical judgment to instantly recognize poor task design.

Core Responsibilities

• Task Execution & Analysis (40%): Run OB-1 against tasks. Analyze results. Understand why it succeeds or fails.

• Task Design Review (40%): Judge if tasks are well-designed, solvable, and test real capability. Spot what's trivial or can be gamed. Refine as needed.

• Curation & Scaling (20%): Filter task batches for quality. Build repeatable curation process as volume scales to 500+.

Required Expertise

• Expert-level understanding of 2+ domains: ML systems, C++ performance optimization, or Verilog/chip design

• IOI/IMO-level competitive programming background (or similar)

• 5+ years building production systems

• 1+ year professional experience with Python and one of Rust or C++

  • Experience with a Typescript a plus

• High bar for quality with ability to articulate why tasks are good or bad

What we offer

• Competitive compensation: $200,000 - $320,000 base salary plus significant equity

• Opportunity to work on cutting-edge AI technology with real-world impact

• Collaborative environment with a world-class team of engineers and researchers

• Access to state-of-the-art computing resources and AI models

• The chance to shape the future of how software is built