Supercharge data flowing into advanced analytics


Robotic data automation paves the way for the next generation of data integration architecture, which is data structure and data mesh.

Every business needs to become a data-driven business, powered by AI and analytics. However, achieving this requires an intelligent and robust data supply chain, in which data travels from its sources in raw form to actionable insights. The tools and technologies that enable this are available, but companies need to reorganize and rebuild around data supply chains.

That’s the word from the recent Robotic Data Automation Fabric and AIOps event, hosted by CloudFabrix, which explored the requirements of a highly functional data supply chain – or pipeline – to support emerging AI and AI initiatives. to analyse. (I had the opportunity to attend the event as a keynote speaker.) “We’re using user data to start making better decisions, like where to allocate resources, where to allocate time, money and investments, how to optimize business, said Sean McDermott, CEO of Windward Consulting Group and another keynote speaker.

See also: Top onboarding challenges today and how to overcome them

The key to achieving a successful data-driven enterprise lies in the combination of two emerging practices: AIOps is the automated and intelligent support of data management, and robotic data automation (RDA) is the data supply chain automation, which supports Ops AI. Although RDA has the potential to underpin the flow of data to AI for all forms and functions, its initial use cases involve computational optimization, with a focus on keeping AIOps working, defining the next challenge on the path to fully automated computing.

“AIOps is also transformational for IT operations,” McDermott said. “We want something smarter when it comes to systems. We don’t want to manually enter data all the time. A client I spoke to was manually updating seven different CRM systems. It makes no sense to do that. We put machine learning and artificial intelligence ahead of IT operations. Conversations we’ve had over the past 20 years include ‘how do I manage alerts? How to do a root cause analysis? How to make a correlation? How to manage all this data?’ The difference now is that we have huge amounts of data, and computing is so much more complex than it was before. The value proposition of AIOps is that it enables digital transformation. The Covid has accelerated a lot of things, and we were moving in the direction of even more mobility and more digitalization of companies.“

AIOps and RDA have become essential as enterprises adopt SaaS, IoT and other technologies, “the infrastructure landscape has become complicated,” said Meenakshi Srinivasan, Global DevSecOps Practice Partner at IBM Consulting, and also a speaker at the conference. “The challenge is to increase manageability.”

Ultimately, AIOps and RDA pave the way for the next generation of data integration architecture, which is data structure and data mesh. “A data fabric architecture stitches together historical and current data across data silos, which hopefully gives you a uniform, unified business view of data,” McDermott said. Data Fabric “breaks data down into reusable components. Data structure is usually associated with digital transformation efforts because it is assumed that data structure architecture will actually reduce costs. Organizations often struggle with managing their data sources and trying to generate value from AI. »

Data mesh, McDermott continued, “is a decentralized architecture, where data architecture is viewed with a high degree of democratization, as well as scalability in mind. It’s really meant to address failure modes currents of traditional centralized data lakes or data platform architecture.”

Transitioning to a data-driven business “won’t happen overnight just because you have a few of the tools in place,” Srinivasan said. “Once we get the base and automation layer well established and start collecting the data. Defining the right datasets, as well as the quality of the data, plays a major role in AIOps. If you don’t have the right data set, this journey will take longer.


Comments are closed.