IT Operations Management has increasingly become a big data problem. The major driver for this is wide adoption of system virtualization which creates a unifying measurement layer across hardware vendors. The result is an explosion in the number of instrumentation points in today’s IT systems. At CloudPhysics, our mission is to use the world’s machine data to take the guesswork out of operations management. We do this by deriving collective intelligence from the continuous machine data sent to us by our customers across the globe. We model how systems will perform by simulating different configurations and analyzing the hidden relationships within them to map the complexities and nuances of IT systems. This allows us to customize big data insights to a specific infrastructure thereby helping drive radical improvements in IT operations. Our back-end data infrastructure and analysis framework handle both semi-structured configuration data and time-series performance data. Cross-user analysis of collected data reveals common patterns in system configuration settings and performance trends. The results enable our users to benchmark against and learn from their peers. Learn about our i) Architecture and data collection/processing techniques for managing 50 billion samples of data per/day using a cluster of Play!Servers powered by Akka Workers backed by clusters of polyglot NoSQL persistent stores. ii) Query Engine that enable administrators to create ad-hoc reports or debug their state of data-center in near to real time I wil share our learnings on the challenges and best practices in leveraging community participation, addressing data security concerns and analysis techniques to extract insight from collected data can be applied more broadly.