The research aims to apply big data in providing an effective model of education in serving the knowledge workers of the municipality. The research method was integrated research (quantitative and qualitative). The components and dimensions of the subject were examined in the form of documentary studies and interviews and identified in the form of educational content with thematic analysis technique. To analyze the qualitative data, the theme analysis method was used using ATLAS TI software, and genetic algorithm and meta-heuristic methods were used in MINITAB software. The research tool (data collection) was the qualitative part of a semi-structured interview with 12 elites, experts, and qualified specialists of Karaj municipality. The sampling method in the qualitative part was non-probability and non-homogeneous purposeful type dependent on the criterion and in the quantitative part, it was simply random. Finally, the proposed model of in-service training for employees was designed and validated. 6 comprehensive themes (planning (comprehensive implementation), learner, teachers, content, educational environment, and infrastructure) were identified in the form of a paradigm model. The results showed that the VIS algorithm had the best performance. Algorithms CNSGA-II and MISA are almost ranked second and have shown almost similar performances. NSGA-II algorithm is ranked next. The NNIA algorithm is in the next position in terms of performance, and the worst performance is assigned to the NRGA algorithm. Organizational innovation based on big data and organizational training improves the performance of knowledge workers and creativity.