Software Engineer – Open Source & SWE-Bench Evaluation

Anyone AI Inc.; United States; Remote; USD 83,000 - 124,000; Part time

Benefits offered by this job

Remote work

Project-based consulting

Flexible hours

Job summary

Компания Anyone AI Inc. ищет опытных Software Engineer для проекта по обзору и оценке задач на основе открытого кода.

Ваша работа будет заключаться в проверке технической выполнимости задач, воспроизводимости окружения и полноты тестов, а также в предоставлении рекомендаций по принятию, улучшению или исключению задач. Требуется опыт работы с реальными репозиториями, PR, тестированием и управлением зависимостями.

Qualifications

• 3+ года профессионального опыта в разработке ПО.
• Опыт работы с большими мультифайловыми кодовыми базами.
• Опыт ревью PR, отладки и поддержки продакшн-кода.
• Сильные знания тестирования и покрытия тестами.
• Опыт управления зависимостями, настройкой окружения и воспроизводимостью.

Responsibilities

• Reviewing coding tasks from real GitHub issues and PRs.
• Assessing clarity and solvability of problem statements and success criteria.
• Evaluating tests for correctness, coverage and robustness.
• Identifying flaky tests, missing dependencies, and environment issues.
• Providing recommendations on acceptance, improvement, or exclusion of tasks.

Skills

Software engineering

Large codebases

Git & GitHub

Unit testing

Dependency management

Reproducibility

Open-source

Python projects

Benchmarks

Docker & CI/CD

Tools

SWE-Bench

Job description

Описание:

Anyone AI is recruiting experienced Software Engineers for a specialized project focused on reviewing and evaluating real-world software engineering tasks derived from open-source repositories. The work involves assessing whether coding tasks based on real GitHub issues and pull requests are technically sound, reproducible, appropriately tested, and representative of the kinds of problems professional software engineers solve every day.

Задачи:

• Real-world bug fixes and feature implementations
• Open-source repositories and pull requests
• Unit tests and test coverage
• Repository setup and dependency management
• Reproducibility and environment configuration
• Task difficulty and complexity
• Multi-file and cross-module code changes
• Technical feedback and quality assessment

You’ll determine whether tasks are clearly specified, technically solvable, supported by sufficient tests, and free from issues such as flaky tests, missing dependencies, ambiguous requirements, or environment-specific behavior.

• Reviewing coding tasks derived from real GitHub issues and pull requests
• Assessing whether problem statements and success criteria are clear and complete
• Evaluating unit tests for correctness, coverage, and robustness
• Identifying flaky tests, missing dependencies, version conflicts, and environment issues
• Determining whether tasks can be reliably reproduced across environments
• Assessing the real-world difficulty and complexity of each task
• Providing clear recommendations on whether tasks should be accepted, improved, or excluded

Engagement Work Type:

Remote Engagement: Part-time, project-based consulting

Focus: Software engineering, open-source code review, testing, and technical evaluation

This role is a strong fit for experienced engineers who enjoy debugging complex codebases, reviewing pull requests, working with open-source software, and evaluating what makes a software engineering problem well designed.

Требования:

• 3+ years of professional software engineering experience
• Strong experience working with large, multi-file codebases
• Experience reviewing pull requests, debugging issues, and maintaining production code
• Strong understanding of unit testing and test coverage
• Ability to evaluate whether tests correctly validate a solution without unnecessarily restricting implementation approaches
• Experience with dependency management, environment setup, and reproducibility
• Strong understanding of Git and GitHub-based development workflows
• Ability to analyze complex technical problems and provide clear written feedback
• Contributions to or maintenance of open-source projects
• Experience with SWE-Bench, SWE-Bench Verified, or similar coding benchmarks
• Experience with major Python open-source projects such as Django, Flask, scikit-learn, SymPy, matplotlib, requests, or pytest
• Experience with Docker, CI/CD, pip, conda, or dependency pinning
• Knowledge of test fixtures, test isolation, or property-based testing
• Experience designing technical assessments or reviewing coding challenges
• Experience with AI/ML evaluation, data curation, RLHF, or benchmark development