In recent years, retrieval augmented generation (RAG) and multi-agent architectures have emerged as promising paradigms across a variety of applications. However, their utilization for automated SEO content generation in the e-commerce sector remains largely underexplored. This study introduces a Multi-Agent Retrieval-Augmented Generation system for automated SEO content creation in e-commerce (ViSEO), a modular architecture designed to stream-line the automation of SEO content generation for e-commerce platforms. ViSEO leverages a multi-agent design, featuring a Supervisor Agent that coordinates specialized task agents for keyword research, data retrieval, and SEO content generation, ensuring a scalable and contextually aware workflow. Using the RAGAs framework, we evaluated three embedding models along with two Large Language Models (LLMs), demonstrating high retrieval precision and contextual relevance, particularly with text-embedding-3-small and gpt-4o-mini. Our findings highlight the framework’s adaptability to multilingual and regional contexts and provide actionable insights into embedding selection, LLM configuration, and multi-agent system construction for robust automated SEO content generation in e-commerce.