About the job
We invite you to submit your CV in English, including your level of English proficiency.
Mindrift expertly connects talented specialists with project-based AI roles at leading technology companies, focusing on the evaluation, testing, and enhancement of AI systems. Note: Engagements are project-based rather than permanent employment.
About the Role
This project is tailored for an experienced Senior Python Developer who possesses extensive expertise in functional testing, alongside robust skills in Linux and Docker. The ideal candidate should be adept at reading code in various programming languages (including C, Rust, and Go) and effectively translating requirements for migration tasks. Additionally, familiarity with tools like Roo Code and Claude Code to facilitate iterative development is essential.
Key Responsibilities
- Develop and execute functional black box tests for extensive codebases across multiple programming languages.
- Establish and manage Docker environments to guarantee 100% reproducible builds and testing across diverse platforms.
- Oversee code coverage metrics and set up automated scoring criteria to adhere to industry-standard benchmarks.
- Utilize LLMs (Roo Code, Claude) to speed up development cycles, automate repetitive tasks, and enhance overall code quality.
Requirements
- Minimum of 5 years of experience as a Software Engineer, predominantly in Python.
- Extensive experience with pytest, including fixtures, session-scoped testing, and timeouts, as well as designing black-box functional tests for CLI tools.
- Advanced Docker skills, including creating reproducible Dockerfiles, managing user contexts, and ensuring secure workspaces.
- Strong proficiency in Linux and Bash scripting, with the capability to debug within containers.
- Familiarity with contemporary Python development tools (e.g., uv, pyproject.toml, packaging).
- Ability to comprehend and analyze code in multiple programming languages with the support of LLMs (such as C, C++, Rust, or Go).
- Experience employing LLMs (Claude Code, Roo Code, Cursor) to enhance iterative development and generate test cases.
- English proficiency at a B2 level or higher.
Preferred Qualifications
- Previous experience with agent evaluation platforms and MCP CLI.
Tools and Technologies: Python (pytest, uv, Pillow), Docker, Bash, Git Submodules, C/C++/Rust/Go (reading), Dagger, GitHub Codespaces, LLMs (Claude Code, Roo Code, Cursor), coverage.py, gcov, kcov.
Benefits
What We Offer
- Freelance, project-based collaboration through the Mindrift platform (powered by Toloka AI).
- Fully remote and flexible engagement—allowing you to choose your working hours and contribution level (20-30 hours per week).
- Compensation for this project varies based on scope and expertise, with AI trainers earning up to $30 per hour.

