Omar
Benkhouya
I build intelligent systems and ship real products. From production RAG pipelines on real banking data, to computer vision at scale, to full-stack web apps.
Projects
Advanced Semantic Search — RAG System
End-to-end Retrieval-Augmented Generation pipeline built on 59 real banking PDFs (~442K chars). Benchmarked 3 retrieval strategies — Dense cosine, Sparse BM25, Hybrid RRF — with a full evaluation suite. Proposed Diversity-Aware retrieval combining MMR with domain-specific banking entity scoring for French-Arabic terminology.
SoccerNet — Player & Event Detection
Computer vision pipeline for broadcast football footage using the SoccerNet benchmark. Player detection, tracking, and event spotting across real match video at scale.
Phone Store — Next.js + Supabase
Production-ready e-commerce platform for mobile phones. Full-stack app with product catalog, authentication, cart, and order management — deployed on Vercel.
Intelligent Robotic Arm — Waste Sorting
Robotic arm system for automated waste classification (plastic, metal, paper). Object detection with YOLOv8, motion control via ROS 2 + MoveIt2, simulation in Gazebo.
Maroc Guide
Web platform for discovering places and travel information across Morocco.
Skills
Experience
AI Engineer Intern — RAG System (PFE)
- ▸Built end-to-end RAG pipeline on 59 internal banking PDFs using BGE-M3 multilingual embeddings and ChromaDB vector store
- ▸Benchmarked Dense, Sparse BM25, and Hybrid RRF retrieval strategies; optimal config achieved HitRate@5 = 90%, MRR = 0.747
- ▸Designed Diversity-Aware retrieval combining MMR with domain-specific banking entity extraction for French-Arabic terminology
- ▸Delivered FastAPI backend and Next.js 14 frontend for internal production use
Full-Stack Developer Intern — iPha Project
- ▸Contributed to the iPha full-stack web application for medical prescription management using Next.js, React, and MongoDB Mongoose
- ▸Integrated AWS cloud services for storage and deployment infrastructure
- ▸Participated in weekly product discussions on Generative AI and Agentic AI integration roadmap
Certifications
IBM
Machine Learning with Python
IBM
Develop Generative AI Applications
IBM
Introduction to Data Analytics
IBM
What is Data Science?
Oracle
Oracle Cloud Infrastructure 2025 Certified AI Foundations Associate
365 Data Science
Certificate of Achievement in SQL
365 Data Science
Deep Learning with TensorFlow 2
365 Data Science
Data Cleaning and Preprocessing with pandas
365 Data Science
Data Preprocessing with NumPy
365 Data Science
Machine Learning with Decision Trees and Random Forests
365 Data Science
Machine Learning with K-Nearest Neighbors
365 Data Science
Machine Learning with Python
Let's build
something.
I'm actively looking for AI Engineering and Full-Stack roles. If you're building something interesting, I'd love to hear about it.