The rapid rise of Generative AI (GenAI) has reshaped the expectations for modern web applications, particularly within Software-as-a-Service (SaaS) products where conversational agents, retrieval-augmented search, and workflow automation are increasingly seen as baseline features rather than experimental add-ons. However, integrating these capabilities into minimum viable products (MVPs) remains a complex challenge. Existing solutions either rely on fragmented microservice ecosystems with high integration overhead, or on traditional web frameworks that lack native support for AI-first functionality. This thesis introduces GenMVP, a Django-based framework designed to streamline the development of AI-native SaaS MVPs. By combining Django’s proven traditional stability with pre-built SaaS subsystems—such as authentication, subscription management, payments, analytics, and content delivery—GenMVP provides developers with a production-ready foundation. Crucially, it embeds generative AI capabilities directly into the framework, offering retrieval-augmented generation (RAG), semantic search, conversational assistance and agentic workflows as configurable modules. The research defines the functional and non-functional requirements of an AI-native SaaS MVP framework and demonstrates how GenMVP satisfies them through modular design, integrated security guardrails, observability, and containerized deployment. Evaluation results confirm that the system achieves both SaaS operational readiness and AI-specific robustness, while case study scenarios illustrate its flexibility across domain-specific applications. By bridging the gap between conventional SaaS development and modern GenAI integration, GenMVP accelerates time-to-market for AI-powered MVPs and provides evidence that traditional architectures remain competitive for next-generation SaaS platforms.