Becoming truly data-driven remains a key goal—and pressure—for organizations seeking to transform its infrastructures. Cultivating a modern, scalable data architecture, capable of delivering continuous value, requires a deep understanding of your unique business goals and the ways in which emerging technologies do—or do not—apply.
John O'Brien, principal advisor and industry analyst, Radiant Advisors, led the annual Data Summit's pre-conference workshop, “Architect Your Enterprise Data Future: Building Strategic Modern Data Platforms,” exploring a proven four-step methodology for translating business goals into architectural realities, assessing various nuances associated with cultivating a modern data infrastructure.
The annual Data Summit conference returned to Boston, May 14-15, 2025, with pre-conference workshops on May 13.
O’Brien explained that at the core of his session is “putting together a plan—an actionable roadmap you can follow [for developing a modern data architecture]—because the number one challenge at companies that I hear from is that they feel like they’re trying to boil the ocean.”
Establishing alignment between each team and the company’s goals is crucial in narrowing enterprise focus. Modernizing data platforms, though technical, impacts each piece of the business, and keeping business priorities in focus will drive that alignment.
Additionally, enterprises must create an architecture that enables people; one that gets people to work faster, avoids creating siloes, and leverages enterprise data. Part of that is eliminating point solutions, where a more integrated data architecture helps increase reusability, quality, consistency, and skills proficiency that, in turn, drives enablement through a sense of seamlessness.
O’Brien then introduced a data strategy framework, which divides alignment into four progressive components:
- Business strategy alignment
- Data and analytics capabilities
- Data platform architecture
- Technology operations
Ultimately, through each of these facets, “We develop architecture to enable a business to do something,” said O’Brien. “We don’t deliver a project—we deliver a capability that allows you to deliver hundreds of projects.”
Delving deeper into each of these components, O’Brien explained that business strategy alignment is key toward driving customer centricity and engagement; accelerating product improvement and delivery times; inviting a data and analytics culture and transformation; and creating innovation cycles and ongoing data product development.
Fundamentally, by identifying the business outcomes you wish to achieve, you inform every part of the architecture as it’s developed going forward, cultivating alignment downstream. Typically, according to O’Brien, enterprises fall into one of four areas of improvement: understanding customer behavior, understanding products usage, increasing operational efficiency, or establishing business model innovation.
Regardless of where your organization falls, becoming a data-driven company is a key part of addressing each of these obstacles. Adopting intuitive tools, optimizing analytics lifecycles, empowering people with self-service, and setting analytics as a strategic priority helps establish a data culture underpinned by data-driven decision making.
Moving onto data and analytics capabilities, O’Brien introduced a comprehensive enterprise capabilities framework, which included analytics delivery capabilities:
- Business intelligence (BI) and reporting: Quickly deliver pre-defined measurements and metrics that track and analyze business progress towards achieving established goals.
- Self-service data analytics: Empower people to easily discover, understand, work with, and collaborate on enterprise data for insights and answering what-if scenarios.
- Data science, machine learning (ML), and AI: Enable data scientists to find and assemble data to build, test, deploy, and manage models that provide recommendations and automation.
- Data product management: Adopt a data delivery methodology to align data teams with consumer priorities while autonomously working within the enterprise data platform.
And data management capabilities:
- Data quality management: Enable business groups to define, manage, and observe quality rules for data sets and elements.
- Reference and master data management (MDM): Enable business groups to define, cleanse, and manage reference data, business entities, and hierarchies for operational systems and analytics.
- Enterprise data governance: Establish an enterprise framework to define roles, responsibilities, and processes to facilitate accountability and stewardship processes.
Architecturally, data architects must contend with a variety of challenges, from addressing industry best practices to business needs for data and analytics solutions and contending with disruptive technologies and vendor products. Ultimately, O’Brien offered these architectural principles:
- Architecture strategy is delivery: Have a vision architecture that is holistic and high-level where pure data architecture projects demonstrate ROI.
- Analytic projects balance risk and impact: Enterprises should balance quick wins and high visibility with risk mitigation and risky projects.
- Embrace delivery for learning: Real data and scenarios will help to surface architecture patterns and decisions; make a data and analytics project work first and then improve.
- Start with pain relief and focused areas: Start small with direct business benefit, focusing on how to deliver business value, while incorporating the time it will take for people to develop cloud competency.
- Adopt a MVP mindset: Architecture projects should be “Minimum Viable Projects,” establishing end-to-end delivery of a reusable architecture pattern at initially a smaller scale.
Finally, regarding technology, O’Brien emphasized that organizations should plan for an open architecture. An architecture that is always evolving, capable of continuously improving with new, appropriate technologies as they emerge enables better, faster, and confident business decisions. With an open approach, technologies tend to be more interoperable with industry standards, which helps consolidate tools and eliminate silos. While specialized technology can still be adopted, limiting the quantity of these tools helps establish a consistent data and analytics structure.
Many Data Summit 2025 presentations are available for review at https://www.dbta.com/datasummit/2025/presentations.aspx.