Ensemble prediction, one of the most rapid developments in numerical weather prediction, has presently become a vital basis for accurate forecasting and assessment of product abundance. In the past three decades, accompanied by the rapid developments in prediction research and techniques, a significant progress has been made in operational technology and systems for ensemble prediction. As the output end of the information facing downstream users in the ensemble prediction chain, the post-processing system has been an integrated platform for the generation of numerous ensemble data, the unification of product-making functions, and the intensification of multilevel forecasting approaches and techniques. In this study, a comprehensive local-to-global review was first conducted for the historical development, current stage, and future direction of the post-processing system and technology for ensemble prediction. Second, the following seven main functions of the post-processing system were summarized: \mathrm{①} Standardized output and distribution of ensemble data; \mathrm{②} Calculation of ensemble mean and spread statistics; \mathrm{③} Analysis of synoptic and climatological diagnostics; \mathrm{④} Generation and issuance of deterministic and stochastic ensemble prediction products; \mathrm{⑤} Extraction and interpretation of big ensemble data and information; \mathrm{⑥} Calibration and improvement of deterministic and stochastic ensemble forecasts; \mathrm{⑦} User-customized product services and visualization. Finally, the unified post-processing system in the China Meteorological Administration-Global Ensemble Prediction System/Regional Ensemble Prediction System (CMA-GEPS/REPS) was discussed in terms of the above main functions. The focus was on finding ways to make better use of the big ensemble data and information from the CMA-GEPS/REPS real-time forecasts to study. Further, the intent was to develop a variety of new ensemble products, particularly including the extreme forecast index, Madden-Julian oscillation, western-Pacific subtropical high, and south-Asian high, as well as learning to apply them to realistic operational forecasting. Overall, the post-processing technique is becoming a predominant research and development direction, building on the advantages of ensemble prediction ranging from high forecast accuracy to actionable insights with significant social, environmental, and economic benefits.