Big data has been around for a decade, yet production big data analysis systems that extract business value are scarce. The problem is complexity at multiple levels, and in 2017 the industry will start addressing those complexities, writes Wikibon Big Data & Analytics Analyst George Gilbert.
Although Gilbert doesn’t expect all the complexity to go away, he expects that leading big data vendors and public cloud business service providers will step up to provide customizable solutions for users. He predicts that:
- Public cloud service providers will drive machine learning applications into the mainstream.
- Machine learning will finally move beyond data scientists to become accessible to mainstream developers.
- While strategic machine learning applications will remain science projects, leaders like IBM, Accenture LLP and Palantir Technologies Inc. will step up to deliver solutions to their customers.
- Live machine learning models with data feedback loops will become the source of sustainable differentiation for enterprises.
Gilbert sees IBM’s creation of application program interfaces based on Watson that developers can import into business applications will be a model for the market. He warns that, in many cases, the first mover in a specific use of shared data will gain an insurmountable lead in specific niche markets.
The example he cites is IBM’s acquisition of The Weather Company, which has developed an air turbulence map for pilots based on live data from their iPads. The Weather Company’s high market penetration in airlines and its first-mover advantage has made it virtually impossible to dislodge from that market with current technology.
In the full Professional Alert, Gilbert examines each of his predictions and their implications for businesses in depth. Wikibon Premium subscribers can read the full analysis here. To become a Wikibon subscriber, look here.