I am an AI Engineer specializing in production-grade LLM systems, retrieval-augmented generation (RAG), and Arabic NLP. I have hands-on experience deploying vLLM inference servers with AWQ quantization (achieving 43x latency reduction), building vector search pipelines with pgvector, integrating Whisper STT, and designing multi-modal AI architectures using Docker and Portainer.
My expertise includes: LLM fine-tuning and optimization, Arabic dialect NLP (MARBERT, Llama), voice agent systems with sub-2s latency, CLIP-based visual classification, and SSE streaming APIs.
I support clients from architecture design to full deployment — whether you need a custom AI assistant, an Arabic voice agent, a semantic search system, or a scalable AI backend. I deliver clean, documented, production-ready solutions tailored to your goals.