Sean Poulley, IBM - IBM Insight 2015 - #ibminsight - #theCUBE
summary
Hadoop consists of 15–20 packages that create real operational complexity, so IBM drives the Open Data Platform to standardize distributions. IBM is investing heavily in Spark as an “analytic operating system” that abstracts across data stores. Real deployments include entity analytics for anti‑money‑laundering and predictive maintenance on Pratt & Whitney engines that forecasts faults so parts are staged for landing. Typical projects spend ~80% of time on data wrangling and one customer plans to hire 500 data scientists.
AI-written summary · about
