Office Hours - Software Engineer- Benchmarking
Requirements
• Solid engineering skills: 4+ years of professional experience building and maintaining complex systems, with strong Python. You write robust, maintainable code and are comfortable diving deep into existing codebases and infrastructure. • Data rigor: Experience preparing, cleaning, and maintaining datasets, and the care to make sure two results are genuinely comparable. • Data rigor: • Comfort with containers and environments: Experience with Docker and building reproducible execution environments. • Comfort with containers and environments: • Collaborative: You work well alongside researchers and scientists and can translate their methodology into working systems. • Collaborative: • Hands-on experience running AI evaluations, or with frameworks like Harbor, Terminal-Bench, or Inspect, is a strong plus. • Evaluations: Python, model APIs, agent/evaluation frameworks, custom evaluation tooling • Models: APIs from the major AI providers, terminal agents, and open-source models via the Hugging Face ecosystem and PyTorch • Environments: Docker • Publishing: React, Next.js, Tailwind • Workflow: GitHub, Slack, Notion, Linear • Experience fine-tuning or post-training open-source LLMs, or other hands-on machine learning work • Experience with agentic, multi-turn, long-context, or tool-use evaluation • Experience validating LLM-as-judge or rubric-based grading setups • Background or strong interest in a scientific or technical domain • Experience building data-heavy dashboards, leaderboards, or visualizations • Open-source contributions or published work related to benchmarks and measurement
Responsibilities
• Prepare and maintain benchmark datasets: Own the data work behind our benchmarks, including cleaning, preparation, conversion into runnable formats, and ongoing maintenance. Validate that tasks are complete, consistent, and executable, and flag ambiguities that would compromise results. • Prepare and maintain benchmark datasets: • Build and maintain evaluation pipelines: Build the infrastructure that runs evaluations consistently across model APIs and terminal agents, so results are reproducible and comparable. • Build and maintain evaluation pipelines: • Build evaluation environments: Create lightweight, containerized environments and viewers for tasking and for evaluating model performance on tool use. • Build evaluation environments: • Support model experiments: Help fine-tune small open-source LLMs and compare baseline against post-training performance. • Support model experiments: • Develop the scoreboard and leaderboard: Build the published views of our results, including model-level, benchmark-level, task-level, domain-level, and rubric-level performance. • Develop the scoreboard and leaderboard: • Build analysis tools: Make it easy to identify recurring failure modes, compare models and agent scaffolds fairly, and track capability improvements and regressions over time. • Build analysis tools: • Collaborate: Work closely with researchers and engineers to make sure evaluation data and outputs are accurate, consistent, and well integrated into what we publish. • Collaborate: • Build tooling for data creation and review: Support expert annotation and data-generation projects by building lightweight HTML viewers and internal tools for task authoring, review, quality control, and structured data collection. • Build tooling for data creation and review:
Benefits
• Medical, dental, and vision coverage • Monthly wellness and fitness stipend • Paid time off policy, along with company holidays • Annual company off-sites (Tahoe, Mendocino, Mexico City, San Diego, Park City) • Parent-friendly policies, remote flexibility, and paid family leave • Pay Transparency Notice • Full-time offers include base salary, equity, and benefits. • This role can be fully remote or hybrid out of our SF or NYC offices. • Don't meet every single requirement? Studies have shown that some candidates, especially underrepresented groups such as women and people of color, are less likely to apply to jobs unless they meet every single qualification. At Office Hours we believe in building a diverse and inclusive workplace, so if you're excited about this role but don't meet every qualification in the job description, we still encourage you to apply. You could still be the right candidate for this or other roles at Office Hours!
Apply in one click
Upload My Resume
Drop here or click to browse · Tap to choose · PDF, DOCX, DOC, RTF, TXT