The U.S. China trade war, Brexit, and COVID-19 have all caused major disruption for all types of businesses. For those without real-time management of their supply chain, the last few months was most likely a roller-coaster of emergencies. That was the focus of a recent webinar presented by RTInsights and TIBCO.
Even businesses with automated systems suffered, as AI systems struggled to comprehend the wild changes in demand due to coronavirus. Some models required manual course correction, and others without constant human oversight may still be experiencing drift.
SEE ALSO: Is Your AI Model Still Accurate After the Coronavirus Pandemic?
“The lessons learned from COVID is purchasing intentions can change overnight,” said RTIsights analyst Joe McKendrick. “Many supply chain systems may not be ready for the rapid shifts in purchasing patterns, without having real-time data telling them what’s happening on the ground, and consistent human engagement.”
In the webinar, TIBCO general manager for analytics, data science, and virtualization, Mark Palmer, talks about implementing a supply chain management nervous system, which is able to continue to function even though a global pandemic.
Businesses collect an overwhelming amount of data, but many are beset by what to do with it. Data warehouses and lakes often turn into silos without proper quality assurance and governance. In a Gartner webinar, data quality was considered one of the key challenges preventing AI adoption.
“For years people have been putting supply chain data into various systems,” said Palmer. “Unfortunately, people have a lot of them and there’s tons of ETL use to move the data around. However, in the last five years, we’ve seen a huge uptick in virtualizing that data. Leave it where it is, create an abstracted view of that data, and let the applications access it. The important part of that it makes you go more quickly, it democratizes access to data, but does it in a way that’s safe and secure and scalable.”
To watch the TIBCO webinar on demand, register here.
.jpg?w=1024)
