Senior Python ML Platform Engineer — XAI Pipeline Hardening

We are looking for a senior Python engineer with strong ML/platform experience to help harden and complete the first stage of our XAI/SHAP pipeline. This is not an exploratory notebook or research-only role. The work is production-oriented: small pull requests, strict scope control, characterization tests, CI gates, Ruff/mypy hygiene, defensive refactoring, and backend/API integration boundaries. The project already has a roadmap and governance. The immediate work focuses on making the existing Python ML/XAI codebase safer, testable, reproducible, and ready for later scientific/model-layer improvements. Initial scope The first stage covers: A1b — inert Ruff autofix / whitespace cleanup T1 — characterization safety net A1c — import hygiene A3′ — SHAP explanation correctness A4 — read-only CLI validate/inspect A7 — legacy quarantine/deprecation A9 — container and CI hardening A10 — backend/API integration boundary Required skills Strong Python engineering pytest and characterization testing GitHub Actions / CI/CD Ruff and mypy `pyproject.toml`, packaging, dependency hygiene Defensive refactoring of legacy code Ability to make small, reviewable PRs pandas, NumPy, scikit-learn Basic ML pipeline understanding Backend/API integration awareness Valuable but not mandatory SHAP / model explainability / XAI MLOps Data contracts Geospatial or agricultural data Google Earth Engine / NDVI / remote sensing Docker Working style We need someone who can work carefully inside a constrained roadmap: no broad rewrites no opportunistic refactors no behavior changes hidden inside cleanup PRs no mixing unrelated roadmap items evidence-based implementation clear local gates before every PR adversarial review mindset The ideal candidate is comfortable saying “this belongs in a follow-up PR” instead of expanding scope. Deliverables Each PR should include: small scoped diff clear description local evidence / gate results tests where appropriate no unrelated changes notes on deferred risks or follow-ups What this is not This is not a notebook-only data science task. This is not a frontend role. This is not a greenfield rewrite. This is not a prompt-engineering role. This is not a broad architecture redesign. We are looking for a senior engineer who can stabilize and complete an existing Python ML/XAI codebase with discipline.