Digital health platforms have emerged as a promising tool for managing chronic diseases in rural populations, particularly diabetes. The study utilised a mixed-methods approach combining quantitative data from electronic health records (EHR) with qualitative insights through interviews. A total of 120 diabetic patients were randomly selected for the EHR analysis, ensuring representation across different regions in Ethiopia. During the six-month period, adherence to prescribed medication was observed at an average rate of 85% among participants using digital health platforms, with significant variability noted between villages (e.g., Village A: 70%, Village B: 92%). The findings suggest that digital health platforms can be effective in improving treatment adherence for diabetic patients in rural settings, though further research is needed to identify contextual factors influencing these results. Future studies should investigate the role of community engagement and local healthcare infrastructure in enhancing platform usage and patient outcomes. Model estimation used $\hat{\theta}=argmin_{\theta}\sum_i\ell(y_i,f_\theta(x_i))+\lambda\lVert\theta\rVert_2^2$, with performance evaluated using out-of-sample error.