Sebastijan-Dominis/hotel_management
This project currently contains a cli-based ml workflow specifically designed for the hospitality industry. It covers everything from the registration of raw data snapshots to model promotion. It also includes a (currently unfinished) small ml service to make it easier to use. It may include more features in the future.
What's novel
This project currently contains a cli-based ml workflow specifically designed for the hospitality industry. It covers everything from the registration of raw data snapshots to model promotion. It also includes a (currently unfinished) small ml service to make it easier to use. It may include more features in the future.
Code Analysis
43 files read · 7 roundsAn end-to-end ML workflow engine for tabular classification/regression with CatBoost, providing reproducible pipelines from raw data registration through hyperparameter search, training, evaluation, promotion, inference, and monitoring with full lineage tracking.
Strengths
Exceptional architectural discipline with protocol-based interfaces, registry patterns, and a generic pipeline runner; the two-phase (broad/narrow) hyperparameter search with resumable state, class-weighting policies, and promotion with file locking demonstrate production-grade thinking for a small-team tool.
Weaknesses
The feature engineering operators are simple (ADR per person, season mapping) and the project is CatBoost-only despite the registry pattern suggesting extensibility; the ML algorithms themselves are standard sklearn/CatBoost wrappers rather than novel contributions.
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
386
Contributors
1
Files
1333
Active weeks
13
Repository
Language
Python
Stars
1
Forks
0
License
MIT