وظائف مهندس جودة البرمجيات في تونس

<h2 class="h5">Job description</h2> <div class="t-break" data-jb-field="description"> <span>Please submit your CV in English and indicate your level of English proficiency.<br> Mindrift connects specialists with project-based AI opportunities for leading tech companies, focused on testing, evaluating, and improving AI systems.<br> Participation is project-based, not permanent employment.<br> About the Role You’ll design coding tasks that challenge frontier AI coding agents.<br> Each task is a self-contained Docker environment with a broken piece of software; an AI agent attempts the fix; automated tests verify the outcome.<br> Your deliverable is the full task package: broken code, tests, instructions, and a reference solution proving the task is solvable.<br> Responsibilities : Invent a realistic developer scenario — a real bug, a broken ETL, a missing feature — not a toy problem.<br> Build a reproducible Docker environment with pinned dependencies.<br> Write a pytest that verifies outcomes, not specific commands — deterministic, non-flaky, and does not leak the fix.<br> Write an instruction.<br>md that reads like a Jira ticket a developer would receive.<br> Write a reference solve.<br>sh proving the task is solvable.<br> Calibrate difficulty so current state-of-the-art agents solve the task 20–60% of the time.<br> Iterate based on feedback from expert QA reviewers.<br> Later: review other authors’ tasks as a QA reviewer.<br> Not in scope Data labeling, prompt engineering.<br> Production code to ship — you design problems and verification for AI agents.<br> Leetcode puzzles — scenarios must look like real developer work.<br> Not every candidate task ships — quality over quantity.<br> Requirements 3+ years of production software development in one backend stack — Python, Go, Node.<br>js, Java, or Rust.<br> Depth in one stack beats breadth.<br> Python + pytest fluency — required regardless of primary stack.<br> The task harness is pytest-based even when the broken app is in another language.<br> Fixtures, parametrize, monkeypatch, timeouts, conftest.<br>py. Docker authoring — reproducible Dockerfiles, pinned dependencies, multi-stage builds when needed, non-root user.<br> Linux &amp; Bash — comfort debugging inside containers (strace, lsof, journalctl); shell beyond set -euo pipefail.<br> AI coding agent experience — Claude Code, Cursor, Roo Code, or similar, on non-trivial work.<br> You can cite a specific time the AI was confidently wrong and how you caught it.<br> English — B2+ written.<br> Not a fit Data Science, ML, or Computer Vision engineers without backend-engineering output.<br> Manual QA testers without automation or test authoring.<br> Frontend-only, low-code / no-code, IT Support, or Business Analysts.<br> Engineers who have never written pytest from scratch.<br> Junior, intern, or assistant as the most recent role.<br> Preferred qualifications Domain depth in Security, System Administration (nginx / systemd / cron), Scientific Computing (NumPy / PyTorch / SciPy), DevOps, or Git internals.<br> Modern Python tooling (uv, poetry, pyproject.<br>toml). Coverage tooling (pytest-cov, coverage.<br>py, gcov, llvm-cov, kcov).<br> Fuzzing or property-based testing (Hypothesis).<br> Prior contribution to agent-evaluation benchmarks or related frameworks.<br> Process Apply → Pass qualification (90-minute sample-task screen + short behavioral interview) → Join a project → Complete tasks → Get paid.<br> Time commitment Onboarding: ~10 hours per first task.<br> Steady state: ~5 hours per task, 2–4 parallel tasks per author.<br> Realistic weekly load: 8–20 hours.<br> Higher volume available for top performers.<br> You choose when and how to contribute; tasks must be submitted by the deadline and meet acceptance criteria.<br> Compensation: Paid contributions, rates up to $35/hour *.<br> Task-based compensation equivalent to hourly rate, depending on performance and volume.<br> Some projects include incentive payments.<br> *Rates vary based on expertise, skills assessment, location, project needs, and other factors.<br> Higher rates may be provided to highly specialized experts.<br> Lower rates may apply during onboarding or non-core project phases.<br> Payment details are shared per project.<br> Apply Submit your CV via the Mindrift platform.<br> Indicate your English level, note this role (Software Engineering Evaluation Specialist — Terminal Bench), and include a GitHub profile link if available.<br></span> </div>