AI Engineer (Multi Agentic Systems, LLM Fine-tuning, Distributed Systems)

About the Role We're looking for an AI Engineer with deep technical expertise in modern NLP and large language models, paired with strong production and infrastructure skills. This is a hands-on role for someone who can take an AI solution from research and experimentation all the way to a reliable, deployed system — and who knows how to use AI tools to accelerate their own work without outsourcing the thinking. What You'll Do Design, fine-tune, and optimize LLMs for production use cases, with a focus on NLP and document-centric workflows. Build and improve end-to-end RAG pipelines, including embedding strategy, chunking, retrieval, and information extraction. Design and build agentic workflows and tool-using systems — function calling, multi-step orchestration, and integration via protocols like MCP. Develop OCR and document-analysis solutions for structured and unstructured documents. Design and deploy agentic workflows with tool use and MCP integrations to automate complex customer processes Deliver multilingual (Arabic/English) NLP solutions tailored to regional customer needs Deploy and serve LLMs in production efficiently, applying optimized inference frameworks and model compression techniques. Own the deployment lifecycle: containerization, orchestration, CI/CD, and ongoing monitoring/maintenance. Collaborate across the team to translate requirements into robust, scalable AI systems. Required Qualifications Strong scientific and technical understanding of AI and machine learning, with particular depth in NLP. Deep, practical knowledge of fine-tuning LLMs and parameter-efficient fine-tuning (PEFT) methods (e.g., LoRA, QLoRA, adapters). Proven, in-depth experience building RAG pipelines end to end — embedding types and selection, chunking strategies, retrieval, and information extraction. Proven experience building agentic workflows and tool-using systems — function calling, MCP, and multi-step orchestration. Hands-on multilingual NLP experience, ideally Arabic/English. Strong expertise in OCR and document analysis. Hands-on experience deploying LLMs to production, including optimized serving frameworks (e.g., vLLM) and quantization techniques. Solid experience with Docker, Docker Compose, Kubernetes, CI/CD, and MLOps practices. Proficiency with Git, server infrastructure, and managing deployments. Familiarity with fine-tuning and inference frameworks/tools such as Unsloth, llama.cpp, and similar. Strong working knowledge of Linux (required). Ability to leverage AI tools to accelerate work — using them to move faster while retaining ownership of the design, judgment, and final result. Nice to Have Experience scaling inference and managing GPU resources cost-effectively. Familiarity with evaluation methods for LLM/RAG systems. Experience with model and data versioning, experiment tracking, and observability for AI systems.