Marketplace Explore 6,000+ AI-powered automations to scale your business.

Employee Engagement & Explainable ML

HR Consulting – Engagement Scoring & Model Explainability

Explainable models identifying employee-engagement drivers across multi-client datasets.

200

Engagement-score questions

Years of multi-client data

66%

Missing values handled

Engagement drivers, explained

From a GLM to XGBoost, LightGBM and Random Forest, explained with SHAP and LIME

Project Details

Client

Qualintra

Sector

HR Consulting – Engagement Scoring & Model Explainability

Year

2023

Key Technologies

XGBoostLightGBMSHAPLIME

Mission

Lead Senior Data Scientist - Employee Engagement Scoring

HR Data Science – Engagement Scoring & Model Explainability

Data Science assignment, Paris / remote.

Pipeline development: machine learning to explain and optimize the HR engagement score (200 questions, 25 years of multi-client data). Migration of the statistical model (GLM): to advanced algorithms (XGBoost, LightGBM, Random Forest). Advanced data cleaning and imputation: (66% missing values), outlier detection, multicollinearity handling (correlation, PCA, regularization). Global and local explainability: SHAP, LIME, sensitivity analysis (Morris, Sobol). Optimization: hyperparameter tuning (Optuna/Hyperopt), business-driven feature selection, cross-validation, custom metrics (RMSE, MAE, R²). Production of dashboards and reports: educational, with personalized HR recommendations.

Contexte & Environnement

Analytics: HR, employee engagement scoring, explanation and prioritization of operational HR levers. Workflows: robust and interpretable ML, clear and actionable reporting.

Team

Lead Senior Data Scientist - Employee Engagement Scoring

Environment

Paris / remote – PySpark, pandas, scikit-learn, PowerBI/Tableau dashboards