The construction of the judicial think-tank big data platform takes standards as the guide, infrastructure as the support, data as the core, security and O&M as the guarantee, and application as the purpose, conducting systematic overall architecture design and implementation. By establishing unified data specifications and standards, it strengthens the collection, management, and application of judicial data, concentrates judicial information resources, establishes a data resource catalog, realizes data reporting, external network push, data exchange, and data sharing, strengthens application system integration and business collaboration, and initially forms a modern trial system covering all businesses and running through the entire process. Through comprehensive access to the court's existing system data and some external data, it builds a data resource capability center and an algorithm capability center, establishes five major data resource capabilities, two major data-type feature extraction capabilities, and AI algorithm capabilities, and implements specific applications for the "more cases than personnel" problem.