<![CDATA[Education holds a central position in the development of citizens and the nation. According to the Ministry of Education and Culture (KEMENDIKBUD), the graduation rate at vocational high schools (SMK) is only 65%. This study was conducted at SMK Ma’arif NU 01 Karangkobar with the objective of implementing an early student graduation prediction system based on academic performance and engagement data. A Streamlit-based Random Forest algorithm was applied to enable schools to anticipate students at risk of not graduating on time and to reduce cases of conditional graduation. The research employed a Research and Development (R&D) methodology with a quantitative descriptive and applied experimental approach. Data were collected through interviews, observations, questionnaires, and literature review. The system was developed using the Waterfall model through the stages of Requirement Analysis, System Design, Implementation, Integration, and Testing. System design involved Data Flow Diagrams (DFD) at Levels 0–3 and Entity Relationship Diagrams (ERD), while the implementation was carried out using the Python programming language, the Streamlit framework, and libraries such as pandas, numpy, and scikit-learn. The user interface was designed using Figma, and diagrams were constructed with Draw.io. Evaluation results based on the Likert scale showed that the system achieved an overall satisfaction rate of 93% from a total of 35 respondents, consisting of 29 student respondents with 93% satisfaction and 6 teacher respondents with 95% satisfaction. This achievement falls into the “very good” category. These findings indicate that the graduation prediction system can effectively support schools in taking anticipatory measures and providing more targeted student guidance, thereby potentially improving the effectiveness of graduation management at the vocational high school (SMK) level.]]>
