Student in the MVA Master at ENS Paris-Saclay and EPITA.

Deep Learning, Foundation Models, Computer Vision, Optimization, NLP, Local LLM integration, Inference systems, Scalable backend services, System design

I am a final-year Computer Engineering student in at EPITA majoring in AI/ML. Also, I am currently working as a student researcher in my school's lab on the adaptation of SOTA speaker verification models to enhance their robustness against audio spoofing attacks and deepfakes.

I am particularly interested by the research field and all the advances related to machine learning and artificial intelligence. I like to keep learning every day. I am open to everything and I love discovering new topics and new people.

Currently working as a research student, my focus is on the adaptation of speaker recognition foundation models to enhance their robustness against audio spoofing attacks.

I’m Samy Yacef, an EPITA MSc engineering student in Data & AI. I focus on building scalable data pipelines and ML-enabled services and have strong SRE/DevOps experience for production deployments. Past roles include a Data Engineer internship at Padoa, medical imaging research, and a prize-winning DevOps co-pilot at a Mistral AI hackathon. At MCP I’ll lead backend/infra or MLOps work and help the team ship quickly and reliably.

MLOps & Production ML Real-time & Streaming Data Recommenders & Personalization Healthcare & Medical Imaging Infra, GitOps & Observability Scalable Backends & Databases Conversational AI & Retrieval

I'm tinkering on NewsBro, a resilient, microservices news recommender: Kafka-based scraping/ingest, Qdrant vector store, TF-IDF & SBERT models, Redis zsets for feed queues, and MLflow/MinIO for artifacts. Deployments use GitOps/Flux on Kubernetes with Terraform, Vault & ExternalSecrets, MetalLB and public/private ingress. The system ingests ~1k articles twice daily and precomputes low-latency recommendations for personalized feeds. Live demo at app.newsbro.c