Processes consume data, processes generate data. Some need a lot of data, like claims or loan, some generate a lot of data like online shopping. And data is generated at a different rates in different processes. Faster the process, faster is the rate at with they need data and generate data, in addition, response times of decisions in the process need to be adjusted accordingly. Some processes need a quick response, like when some highly caffeinated trader presses B instead of M while trading and takes the whole market down or algorithmic trading with a mind of its own. Some processes need continuous feed of data from many sources for continuous course corrections, like managed health care.
In any process, we make decisions along the way. The decisions are based on rules that we have learned and put in place, information generated from historical data, data currently being generated and also events and event patterns emerging. Depending on the type of decision and when that decision is being made in the process we either need to rely on rules, predictive analytics, fast data analytics or complex event processing.
Depending on the required response times and customer expectation we need to utilize these technologies at our disposal to make right decisions at the right time with as much certainty as possible. Whether it is about an up-sell opportunity during an online sales process or it is a fraud management process or a loan origination process, delivering results at the right speed is critical.