Clement Soens

💼 Professional experience

Agilytic

Data Analyst/Scientist Since 2021

• Advertisement: Developed a full pipeline for data ingestion and standardization on a daily basis, deployed using SQL, Spark, and Databricks. Creation of a data model and a semantic model in PowerBI.
• Retail Forecasting: Led a team of four on a 7-month forecasting project for a multinational firm (€12bn+ annual revenue), delivering an ML-based solution that improved existing sales forecast accuracy by 15% and deployed in the Azure Environment.
• Retail Analytics: Collected and formalized business requirements. Modeled customer behavior and assessed marketing campaign performance using causal ML, A/B testing, and propensity matching with LGBM and XGBoost.
• CRM Automation (Public Sector): Elaborated functional requirements and created a data cleaning automation tool for CRM systems, integrating cloud infrastructure and PowerBI reporting.
• Business Model Impact Analysis: Estimated the impact of a business model change using SQL & Python analysis.
• Project Management: Led a team of four on a five-month forecasting project for a CAC-40 company (€12bn+ annual revenue), delivering actionable predictive solutions.
• Media Forecasting: Developed an ML-based sales forecasting tool, improving the accuracy of existing methods by 15%.
• Predictive Maintenance: Analyzed processes and flagged outliers to improve equipment reliability. (Python-based analysis)
• Customer Retention (Retail): Developed predictive models for estimating Net Promotion Scores using XGBoost.
• Customer Segmentation (Public Sector): Conducted customer segmentation and reporting using PCA and KNN analysis.
• Marketing Campaign Evaluation (Non-Profit): Applied causal inference to assess marketing effectiveness with propensity matching and sensitivity analysis.
• Fraud Detection (Banking): Built fraud detection algorithms for salary slips using logistic regression, decision trees, SVM, and random forests.
• Invoice Data Automation (Banking): Implemented NLP and image processing for automated invoice data extraction.
• Debt Recovery Prediction: Developed an SVM model to predict debt recovery likelihood, with SQL integration.
• Recruitment Automation: Used NLP for automated extraction of skills and experience from resumes.

Université Libre de Bruxelles (ULB)

PhD Candidate & Teaching Assistant in Innovation Valuation Since 2022

My academic focus is on the use of Advanced Econometric and new latest AI techniques to assess the monetary (and social) value of patent granting. I use NLP analysis to perform feature engineering on patent data; currently exploring various regressors to assess the environmental value of innovation.

I teach third-year students the econometric models used to perform a monetary valuation of IP rights.

Yves Rocher

Intern 2019 Management Accounting in the Trade Logistics department.

🦾 Certifications, trainings

Methodologies

• Econometrics and Time Series Analysis: Research in econometrics with techniques such as ARIMA, SARIMAX, Griliche’s model, and Difference-in-Differences (Diff-in-Diff).
• Parametric and Nonparametric Estimation and Testing: Application of both parametric and nonparametric approaches for statistical estimation and hypothesis testing.
• Regression and Classification: Expertise in both linear and non-linear methods for regression and classification, including SVM, XGBoost, and LGBM.
• Data Modeling: Development and validation of complex data models for predictive analytics and decision support.
• Descriptive and Predictive Analytics: Conducting both descriptive and predictive analyses to extract insights and forecast future trends.
• Machine Learning Techniques: Proficient in implementing advanced machine learning algorithms, including causal inference models, A/B testing, and propensity matching.
• Optical Character Recognition (OCR): Implementation of OCR for text extraction from scanned documents and images.
• Image Processing: Utilization of image processing techniques for data extraction, automation, and feature enhancement.
• Causal Inference: Application of causal ML and inference methods, including propensity score matching, sensitivity analysis, and experimental design (e.g., A/B testing).

Software & programming

• Python, Matlab, R
• Timi Suite
• Power BI
• Databricks
• Google Looker
• Arduino

🎓 Academic credentials

• M.Sc. Research in Statistics and Econometrics European Center for Advanced Research in Economics and Statistics (ULB), 2024 Cum Laude
• M.Sc. Business Engineering Solvay Brussels School of Economics and Management, 2021 Cum Laude
• Quantitative Technics for Economics and Management (QTEM), 2021

🇺🇳 Languages

🇫🇷 French: Native

🇬🇧 English: Fluent

🇳🇱 Dutch: Working proficiency