As the cost decrease of hardware and software with proven technologies, the traditional way to analyze and manage data has obvious shortages to meet the needs of business development. Because the monitoring and evaluation data of environmental protection matches the characteristics of big data, it is possible to build a new data management platform using big data technology. In this study, we design the three-layer architecture and physical implementation proposal of big data management platform for environmental evaluation, with the research of elastic expansion, stream processing, data lake, massive parallel processing, machine learning, and other technologies. We try to resolve the data processing issues of high speed, deep insight, and intelligence so that to provide high efficient support to business management.