Senior Agentic AI Engineer Hello At MBC, we are continuing to grow, and we are now looking for a Senior Agentic AI Engineer to join an innovative project focused on Agentic AI, multi-agent systems, advanced machine learning, and intelligent engineering solutions. We are looking for a highly skilled professional with strong expertise in AI architecture, LLMs, Python, C++, and production-grade AI systems, who is passionate about building scalable, reliable, and cutting-edge AI solutions. About MBC May Business Consulting is an international consulting firm providing advisory services in digital transformation, change management, efficiency, and control. We take a fully client-centric approach, proactively engaging in every project to maximize the value we deliver. We have offices in Spain and the United Kingdom. We work on international projects with financial institutions, fintech companies, and startups.

About the Role
We are looking for a Senior Agentic AI Engineer with extensive experience designing, developing, and deploying production-ready AI agents and multi-agent architectures. You will play a key role in building intelligent systems that combine LLMs, advanced reasoning, model orchestration, retrieval-augmented generation (RAG), and engineering intelligence. The idóneo candidate has a strong software engineering background, hands-on experience with LangGraph, agent communication protocols, model training, and MLOps, and the ability to transform complex technical requirements into robust AI solutions. Responsibilities Agentic AI Architecture & Multi-Agent Systems Design, develop, and deploy multi-agent orchestration architectures using LangGraph, including planning, routing, reasoning loops, state management, and human-in-the-loop approvals. Build Mixture-of-Agents (MoA)



workflows to coordinate specialized agents and improve decision-making. Implement Mixture-of-Experts (MoE) and Mixture-of-Thoughts (MoT) strategies to optimize model performance, latency, and cost. Develop secure agent interoperability using MCP, A2A, and ACP protocols. Implement agent memory, context management, tool calling, approval workflows, and traceability. LLMs, RAG & Model Development Build advanced RAG pipelines, including document extraction, embeddings, vectorization, and hybrid search. Work with vector databases such as pgvector, Qdrant, Weaviate, or Milvus. Fine-tune and deploy open-weight language models using LoRA, QLoRA, PyTorch, and distributed training frameworks. Develop and optimize Small Language Models (SLMs) for efficient cloud, on-premise, and edge deployment. Apply model training techniques including supervised fine-tuning, RLHF, DPO, distillation, and quantization. Explore multimodal AI solutions combining text, images, 3D data, and engineering documentation. Engineering Intelligence & Simulation Integrate AI agents with CAD/BIM data, digital twins, and simulation environments. Develop hybrid AI solutions combining rules-based reasoning and machine learning. Implement optimization algorithms, graph search, constraint solving, and explainable decision-making. Explore world foundation models, including NVIDIA Cosmos, for simulation, synthetic data generation, and physical reasoning. MLOps,



AgentOps & Deployment Build and maintain scalable MLOps and AgentOps pipelines, including CI/CD, versioning, monitoring, and model evaluation. Implement observability and tracing using LangSmith, Langfuse, or similar tools. Establish AI guardrails, reliability mechanisms, and performance evaluation frameworks. Deploy secure AI solutions using Docker, Kubernetes, Linux, GPU cloud infrastructure, and private on-premise environments. Collaborate with multidisciplinary engineering teams to deliver reliable, production-ready AI solutions. Requirements Master's degree or PhD in Computer Science, Artificial Intelligence, Machine Learning, Applied Mathematics, Computational Engineering, or a related field. 6 years of software engineering experience, including at least 3 years developing production ML, NLP, or LLM systems. Expert-level proficiency in Python and strong experience with modern C++ (C++17 or later). Proven hands-on experience with LangGraph and production multi-agent systems. Strong knowledge of MCP, A2A, and ACP protocols and secure agent communication. Advanced experience with RAG, embeddings, vector databases, and hybrid retrieval systems. Solid understanding of MoE, MoA, and MoT architectures. Practical experience with PyTorch, model fine-tuning, LoRA/QLoRA, RLHF/DPO, and distributed training. Strong understanding of SLMs, multimodal models, and agentic AI runtime architectures. Experience developing rules-based or hybrid AI/ML systems. Strong background in MLOps, Docker, Kubernetes, CI/CD, model serving, and monitoring. Knowledge of REST APIs, gRPC, networking, and secure cloud/on-premise deployments. Familiarity with digital twins, simulation, and world foundation models such as NVIDIA Cosmos. Fluent English, both

📌 Senior Agentic AI Engineer (España)
🏢 Izertis
📍 España