Improving the level of enterprise internal audit is the focus of reducing the business risk of enterprises. This paper is based on the hierarchical analysis method to mine the indicators affecting the quality of enterprise internal audit work, and construct the indicator system. Combined with big data cloud computing and K-Means algorithm, it carries out clustering analysis and character portrait of enterprise employee behavior data to find the types that need to be focused on by auditors. Take procurement approval behavior as an example to study the specific operation of internal audit quality optimization based on cloud computing. Compare the scores before and after the digital transformation of internal auditing to determine its quality improvement effect. The results show that through cloud computing statistics and cluster analysis, it is found that there is procurement fraud within the enterprise where the unit price of procurement is higher than the market price by 400 yuan, and it is necessary to focus on rectifying the internal auditing process in all aspects of procurement. The score after the digital transformation of internal audit is 81.25 points, much higher than the 63.38 points before the transformation, and the audit quality is improved greatly.
