Principal Agentic AI Engineer & Harness Engineering Lead

Role overview

Mandatory experience

• 12+ years of software engineering, solution architecture, or enterprise technology experience.
• 5+ years of hands-on experience in AI, Machine Learning, Generative AI, or Agentic AI solutions.
• Demonstrated delivery of multiple production-grade AI projects from design through deployment.
• Strong experience training, mentoring, and leading AI engineering teams.
• Experience designing enterprise-scale AI platforms and an AI Center of Excellence.

Agentic AI and development tools

• Expertise in autonomous agents, multi-agent architectures, planning, reasoning, tool/function calling, memory, orchestration, and evaluation.
• Relevant areas include human-in-the-loop systems, agent governance and monitoring, Model Context Protocol (MCP), and Agent-to-Agent (A2A) communication.
• Listed development platforms include Claude Code, GitHub Copilot, Cursor, Windsurf, ChatGPT, Microsoft Copilot, Replit, Bolt.new, and Lovable.
• Listed agent frameworks include LangChain, LangGraph, CrewAI, AutoGen, Semantic Kernel, OpenAI Agents SDK, PydanticAI, and LlamaIndex.

Harness engineering and deployment

• Work includes AI development accelerators, prompt engineering frameworks, agent development kits, reusable components, testing harnesses, and evaluation frameworks.
• Relevant governance areas include development standards, AI security frameworks, and enterprise delivery methodologies.
• Deployment areas include on-premise and air-gapped AI, GPU infrastructure, multi-GPU inference optimization, infrastructure sizing, and edge AI.
• Listed tools include Ollama, vLLM, TensorRT-LLM, LM Studio, Open WebUI, and Hugging Face Transformers.

Model optimization and enterprise search

• Fine-tuning approaches include LoRA, QLoRA, and PEFT; quantization includes INT8, INT4, GPTQ, AWQ, and GGUF.
• Other areas include inference and AI cost optimization, context engineering, tokenomics, and performance benchmarking.
• RAG areas include enterprise, graph, agentic, and hybrid search, knowledge graphs, and semantic search.
• Listed vector search tools include Pinecone, Weaviate, Milvus, Qdrant, and Azure AI Search.

Pay

• From AED 20,000 per month.

Workplace

• Work location is in person.

Responsibilities

• Architect, develop, deploy, optimize, and govern enterprise-scale Agentic AI solutions.
• Design and deliver production-grade AI applications, autonomous agents, multi-agent systems, AI copilots, enterprise AI platforms, and AI engineering frameworks.
• Lead AI engineering teams and mentor and enable engineers in modern AI engineering practices.
• Design agent planning, reasoning, tool and function calling, memory, orchestration, evaluation, human-in-the-loop, governance, and monitoring capabilities.
• Develop AI accelerators, prompt engineering frameworks, reusable components, testing harnesses, and evaluation frameworks.
• Establish development standards, AI security frameworks, governance, and enterprise delivery methodologies.
• Design and optimize on-premise, air-gapped, GPU, multi-GPU inference, and edge AI deployments.
• Optimize models through fine-tuning, quantization, inference optimization, and performance benchmarking.
• Develop enterprise RAG, graph RAG, agentic RAG, hybrid search, knowledge graph, and semantic search capabilities.
• Support AI cost optimization, context engineering, and tokenomics.