LinkedIn uses Galene ( https://engineering.linkedin.com/search/did-you-mean-galene ) infrastructure to create and manage search indexes which gives our users best and most relevant search experiences. Building indexes is always a pain when your user base is more than 450+ million members. It was also very tough for us. Building and deploying indexes took more than a day for a engineer with so much of baby sitting whole day during that period. This also requires more than 12 steps to create on hadoop, verify on index built , product build, verify again in Early integration testing and then deploy. I have automated all the steps in Indexmanager project using python and flask . Now indexes can be scheduled, started and managed from single place and everything is automated.