👋 konnichiwa, I'm
Rayen Dhahri
I build efficient AI — optimizing neural networks through sparsification, quantization, and distillation for real-world deployment on constrained hardware. Published at NeurIPS, worked with Zeiss, BMW, Huawei, Intel, and more.
Model Compression
Bayesian Deep Learning
On-Device AI
Highlights
What I Do
Neural Network Optimization
Sparsification, pruning, and compression of deep learning models for efficient deployment.
LLM Efficiency
Making large language models faster and lighter through quantization, distillation, and structured sparsity.
Edge AI Deployment
Deploying optimized models on constrained devices — Jetson, FPGAs, microcontrollers, and more.
Research & Publications
Published at NeurIPS '24 & '25, AABI — Bayesian sparsification, Quant-Trim, and scalable sparse models.
Blog
Latest Posts
Thoughts on efficient AI, research notes, and engineering lessons.