Rajae Yamoul

Designed and deployed end-to-end AI systems used in real operational workflows for payroll and HR
processes
Built OCR and document understanding pipelines combining layout analysis, OCR engines, Camelot,
Tabula, and object detection models (YOLO, SpaCy, Docling)
Transformed unstructured documents (payslips, employment contracts, collective labor agreements)
into reliable production-ready data pipelines
Owned the full AI lifecycle: preprocessing, model training, validation, deployment, monitoring, and
continuous improvement
Worked closely with non-technical stakeholders (payroll managers, HR teams) to validate outputs in real
world conditions
Monitored data drift, model errors, and operational failures, improving robustness and reliability of
deployed systems

Worked on NLP-based text analysis for ESG data collection across African markets
Implemented text preprocessing pipelines: tokenization, lemmatization, vectorization
Applied classification, sentiment analysis, and topic modeling using transformer-based models
Evaluated linguistic outputs using embeddings, similarity measures, and classification metrics
Supported data labeling, validation, and quality control processes for AI training datasets