Steve Dine is a BI and enterprise data consultant and industry thought leader who has extensive experience designing, delivering, and managing highly scalable and maintainable modern data architecture solutions.

Steve combines a strong business acumen with hands-on technical experience and is committed to providing value by helping companies transform their data assets into a competitive advantage. His experience spans Healthcare, Manufacturing, Financial Services, Life Sciences, and Retail. He is currently President at Datasource Consulting.

Key Takeaways

• Traditional architectures are not meant to handle today's large data volumes
• Modern data architectures are distributed, include multiple platforms, support analytic and operation workloads, and typically involve cloud
• Measure your workload needs and match it to a capable platform
• Not all platforms need to be distributed
• Data virtualization tools are rising in popularity, but it's not a magic elixir
• Managing platforms, and often thousands of views, especially when data changes, is a huge challenge
• Don't modernize without a business case, don't let it become a pure IT project

Below is one question and answer from the podcast

Wayne Eckerson: What are the Hallmarks of a modern data architecture?

Steve Dine: Modern data architecture generally includes multiple platforms. They include both relational and non-relational technologies. Secondly, it’s generally architected in a distributed nature, so we’re leveraging platforms with multiple nodes, and data is spread and processed across those nodes. Generally, there’s some level of distributed architecture that may or may not involve Cloud. In a lot of cases, people will assimilate modern architectures with Cloud, but you can have a modern architecture and not be in the Cloud, per say. Lastly, in my view, it supports both analytic and operational type workloads.

We’ve started to architect and collect requirements around enterprise data requirements, not just the data warehouse. I consider the data warehouse to be like the shovel at the end of the parade. It collects all the data from all the systems and then we try to manage it at the end of its cycle. From my perspective, to be successful today you need to manage your data across the entire enterprise. We’re looking at modern architectures; at how we actually design so we can support the data management requirements across the entire organization, whether that’s operational requirements within operational systems or data that needs to be moved or potentially integrated into an operational system. MDM is a great example of a component of modern data architecture. But in today’s world, we need to ensure the data is integrated across all the different operational systems if we want to move at real time or the speed of business.

We’re addressing requirements from an enterprise perspective.  So we have to support both the operational analytic needs, and a lot of our customers have BI organizations that are being transformed from just managing the analytics to supporting these operational applications, whether they’re web-based applications or they’re on-prem non-web-based applications. Those have different data management characteristics and requirements with regard to how we support them. Our modern architectures need to support not just the data warehouse and the analytic requirements, but they also need to support the operational applications.