Practical Considerations for Real-Time Business Intelligence
Real-time business intelligence (BI) must satisfy two requirements to be effective in industry:
(1) "Time is money": it costs money to reduce latency. The decisions based on low latency must justify the investment.
(2) "Data must be actionable": the information, set within the correct context, must be provided to the right people at the right time so that an effective business decision can be made.
In this talk I examine some of the history of BI in industry and the methods used on the path to the nirvana of real-time BI.
Schneider has nearly twenty years experience in data processing and business
intelligence, spanning research and industry. He currently is a Data Architect
at Yahoo! where he is grappling with storage and query challenges with
peta bytes of data. Prior to joining Yahoo!, Donovan was a Senior Architect
and Director of Engineering at Siebel Systems for the Siebel Analytics
product - a federated, business intelligence platform at the center of
Siebel's suite of analytical applications. Donovan architected the query
execution engine in the Red Brick data warehouse, and performed query
processing research while a