Large-scale automatic detection of individual animal phenotypes is a long-standing desire in animal breeding. Further-more, the recent advances in AI, coupled with the availability of technologies such as RFIDs, Wireless Sensor Networks, and Camera Systems, have made it possible to automate the process of animal phenotype detection. An automated phenotype detection process has several phases, comprising of data collection, data fusion, data storage, data analysis/AI model development, and data serving. To facilitate the automated phenotype detection process, we developed a data analytics platform architecture. The designed architecture proposes various approaches vis-a-vis several phases of the automated phenotype detection process, along with implementing these approaches on Dutch nationalHPC infrastructure. Also, how data analysis results are servedto the various end users is very important for the success of adata analytics platform. To this end, we propose a scalable data visualization pipeline that enables animal behavior scientists to deduce quick summaries about animal behavior on a farm.