Real-Time Analytics and Operational Decision Intelligence

Real-time analytics is the processing of data as it becomes available and usually with minimal latency. Such analytics finds applications in diverse domains, including manufacturing, healthcare, transportation, finance, news media, sports, and telecommunications. A common requirement across these scenarios is the need for complex decision-making integrated into the real-time analysis process. Such analytics thus extends beyond traditional monitoring, alerting, and dashboarding toward automated and autonomous Operational Decision Intelligence (ODI).

Five trends help propel operational decision intelligence into the mainstream of real-time analytics: (1) data stream requirements in response to Service Level Agreements (SLAs) from business process applications; (2) automatic or semi-automatic decision models to process the data streams; (3) causal inference to better understand the processes being monitored; (4) explainability techniques such as counterfactuals to foster some level of trust in the results of automated decisions; and (5) integration of external operational constraints such as available inventory, and external signals such as market pricing or actions by competitors. Note that SLAs condition response times for the analytics, not for input latency.