Software reliability prediction is crucial in software engineering. Existing models often overlook valuable data from the software process. This study introduces a new approach using a software process database to enhance reliability prediction. By analyzing software process data and designing measurement units, a more accurate prediction model was developed. This model aims to improve the credibility and accuracy of software reliability predictions, paving the way for further research in the field.