Claude Code plugin for Kaggle competitions: turns Claude into a Grandmaster-level AI teammate with leak-free cross-validation, experiment tracking, tabular/vision/NLP/LLM/time-series playbooks, ensembling, code-competition kernels and final-submission selection.
CRISP-DM data mining and machine learning workflow for Jupyter, JupyterLab, Colab, VS Code and Claude Code, from problem framing to a deployed, explainable prediction service.
Hypothesis-driven data science skills: sharpen a hypothesis, write the experiment contract, slice it into a pipeline, implement it expectation-first in notebooks, review it, and turn results into an article.
Inventory modeling, data import, LCA calculations, contribution analysis and graph traversal, prospective databases with premise, time-explicit LCA with bw_timex, and life cycle optimization with optimex.
Bayesian analysis in R with brms and Stan, with its record and its write-up: a log of each model, check and decision, a workflow record generated from it, the methods and results sections of an article in one file, and every number in the draft checked against the saved output.