Streaming data analysis is an additional stage of analytics that insurance & other sectors should not neglect to have a complete picture of customer experience and CRM.
Take telematics which already offers new information about driver behaviour. But is is only partial information. That aggressive braking might be to avoid a child running onto the road, a car driving out from a side-turning. The insurance company can, of course, request the driver to explain the situation but that is costly, time consuming and manual.
Dash-cams can stream the data to offer a full-insight. Machine-learning technologies like IDOL from HPE can filter out bad driving from acceptable avoidance manoeuvres.
For insurers that have already digitally transformed their business Analytics platforms like Logi Info or Tibco's Spotfire can analyse streaming data in purpose built applications. For those that have not, new insurance platforms like 360Globalnet digitally transform the whole claims management process putting the claimant at the heart of the process. This without disrupting legacy systems.
A key added benefit of these platforms is that as claimants gain from self-service claims processing the insurer has real-time and complete analysis of each and every claim from initial claim form to repair or replacement- bi-directional data streaming!
The insurer digitises "Command & Control Centres" including Call Centres. A logical next stage is to add location intelligence i.e. join location, customer, claim and enterprise data to gain new insights for better decisions and predictions.
In this case the insurer gains from information streaming in data rivers from work processes and the historical data in data lakes. This is complete data - a better term than big data.
This is not a panacea for insurers that refuse to address the issue of digital transformation. Those that insist on retaining large fiefdoms like Risk Management & Fraud. Only when these are digitised should they look to embrace streaming data. For a longer look at this topic see "The Future of Streaming Data" by Mark Palmer
That is, if an automated driver feature in your car causes an accident, who is at fault? The Tesla software engineer who wrote the bug in the software? The “driver” who didn’t touch the wheel? The driver of the vehicle that may be have been hit by the autonomic car, but was also partially at fault? And, relatedly, what is the obligation of all insurance companies to gather and analyze massive amounts of streaming forensic data from insured, automated assets in order to decide? How can risk be mitigated in real-time by monitoring streaming data?