AI Architect

About the role

Capgemini Engineering is hiring an AI Architect in Dubai. You will own the architecture and industrialization of AI capabilities across Generative AI, Agentic AI, Computer Vision, local and cloud models and AI-enabled business processes. You will act as the link between business requirements, data, models, infrastructure, security and operations, making sure every AI solution is measurable, governed, secure and production-ready.

Key responsibilities

• Define target architectures for LLM, RAG, Agentic AI, Computer Vision and multimodal capabilities, including agent design, memory, function calling, human-in-the-loop controls and multi-agent patterns.
• Architect and implement AI and business-process orchestration solutions using platforms such as Langflow and Flowable, turning business processes into workflows that integrate people, AI agents, APIs and enterprise systems using BPMN 2.0.
• Establish and govern the AI evaluation strategy, including Golden Dataset methodology, task-specific metrics, benchmarking, regression evaluation and separation of training and evaluation data.
• Design production observability frameworks covering drift detection, quality degradation monitoring, feedback loops and continuous re-evaluation of approved AI solutions.
• Design local and sovereign model architectures using Ollama, vLLM and the Hugging Face ecosystem, weighing quality, security, sovereignty, cost and operational trade-offs.
• Define RAG pipelines (ingestion, chunking, embedding, vector search, hybrid retrieval, reranking) and Computer Vision pipelines (classification, detection, segmentation, anomaly detection).
• Govern the full MLOps and LLMOps lifecycle, embedding security, Responsible AI principles and architectural governance.
• Act as the AI architecture authority within the solution, working with the Lead Solution Architect and guiding AI/ML and Computer Vision engineers.

Required profile

• Proven experience as an AI Architect or senior technical architect with 8-12 years of hands-on experience in GenAI, LLM, RAG and Agentic AI.
• Practical use of IAF, TOGAF or an equivalent recognized architecture framework.
• Demonstrated ability to define and run AI evaluation strategies, including Golden Dataset design, benchmarking, drift detection and observability.
• Experience architecting AI and business-process orchestration solutions with tools such as Langflow, LangChain, LangGraph or Flowable, and a solid understanding of BPMN 2.0.
• Strong stakeholder communication skills and experience in security and Responsible AI by design.

Required skills

• Python, REST APIs, vector databases, GPU/inference concepts, containerization and cloud AI platforms
• MLOps and LLMOps governance
• Langflow, LangChain, LangGraph, Flowable, Ollama, vLLM, Hugging Face

What we offer

• Dedicated innovation labs focused on 5G, 6G, AI, Autonomous Vehicles and Quantum.
• A digital campus and academy with more than 60,000 hours of training and digital learning on GenAI, data and sustainability.
• Engineering Horizons, a global conference for Capgemini Engineering colleagues.