Raed Bourouis

I’m a Computer Science graduate building production-ready AI systems, machine learning models, and data pipelines. I focus on turning raw, messy inputs into reliable outputs through solid ingestion, orchestration, and monitoring.

In my AI & Data Engineering internship at Talan Consulting, I built an agentic case-tracking system that reconstructs unified state for concurrent dispute cases across 16+ RPA robots using semantic interpretation of 120k+ logs/day. I designed cross-robot entity resolution and a log-ingestion pipeline to unify heterogeneous identifiers into coherent, auditable case timelines.

I also engineered a LangGraph reasoning engine that classifies workflow stages and flags anomalies across 500+ cases/day, replacing manual triage. To make the system usable, I shipped case-exploration and anomaly-review dashboards adopted by the Smart Automation team.

Previously at Union Internationale de Banques, I engineered an end-to-end document processing AI system, reducing manual data entry by 85%. I implemented a PaddleOCR pipeline for field detection and text extraction, then built the full OCR → structured extraction → PostgreSQL → dashboard flow, plus a predictive model to flag high-risk submissions for manual review.