Building smart cities represents a national priority for Indonesia to enhance global competitiveness, with artificial intelligence (AI) driven smart government as a key enabler. However, implementation faces significant challenges including unmeasured organizational maturity, lack of service innovation, fragmented governance, and minimal citizen engagement leading to government institutions' failure in achieving smart city vision. This study aims to develop a holistic conceptual model to identify critical success factors and evaluate processes that integrate public services, fostering AI-driven smart government innovation at strategic level. This research employs mixed-methods exploratory sequential design combining qualitative techniques (Systematic Literature Review, expert interviews) with quantitative validation (citizen survey, statistical analysis). The model was constructed using Factor Analysis, Thematic Analysis, TOGAF framework, and multidimensional view with validation through triangulation, expert judgment, Focus Group Discussions, and statistical analysis. Results show a comprehensive model consisting of 6 dimensions, 17 key components, and 5-layer organizational architecture with high reliability (Cronbach's Alpha 0.709-0.866) and expert consensus (86% agreement in Fuzzy Delphi Method analysis). This framework, referred to as SEMAR v1.0 (Smart Government Nusantara), serves as a benchmark for assessing the maturity and readiness of local government institutions in Indonesia. It offers the potential to improve SPBE scores through systematic evaluation, while also providing a theoretical foundation for smart government scholarship and a practical blueprint for policy implementation.