In this talk, we present a decentralized industrial analytics platform based on the PI System. The design of our platform was shaped by extensive interviews with industry veterans and experts which yielded major pain points like cost, latency, manpower requirements and privacy that plague industrial analytics frameworks of today. With the help of state of the art research in decentralized computing and industrial analytics, we have developed a novel and unique solution that alleviates a majority of pain points associated with industrial analytics. Our framework leverages a historian and Edge Data Store services in order to drastically reduce the total cost of ownership leading to lower latencies, faster decision times and complete data privacy as well.