The space of analytics is one that evolves in tandem with the data needs of any organization. With the modernization of data infrastructures and data architectures comes advanced analytics needs, ranging from going real-time to enabling AI and ML use cases, self-service, and improved performance and scalability. While the needs are varying, the accompanying challenges of modern analytics are equally as diverse.
Experts in analytics joined DBTA’s webinar, The Future of Analytics: Data Warehouses, Data Lakehouses, and Data Mesh, to offer IT leaders and practitioners best practices and new technologies for navigating where modernity meets analytics.
Paul Lewis, chief technology officer at Pythian, framed his discussion of analytics around the ever-popular sphere of generative AI (GenAI).
Lewis explained that GenAI “will have many beneficial applications in a variety of industries. However, organizations will need to address privacy, security, and compliance risks before they leverage this technology and access the next step in their analytics journey.”
Fortunately, Pythian designed the following guide for enabling modern data analytics and unlocking the “promise” of data, particularly as it relates to a GenAI use case:
- Design a data platform to support your data strategy including data warehouses, data lakes, and data mesh.
- Build your governance programs to enable your data lake.
- Ensure every feature and investment has a vocal business champion.
- Make it modular.
- Avoid analysis paralysis when picking data lake components.
- Decouple process steps from technical implementations.
- Match your transformation patterns to your flexibility and operational needs.
- Metadata—store it, manage it, govern it, and use it!
- Deploy AI/MLOps capability adjacent to data platforms.
- Data Quality = User Trust = Data Adoption
In addition to this list, Lewis explained that “whether it’s a data warehouse, data lake, or data mesh that you draw on to manage your data, the state of your data governance is a crucial part of moving toward a future where GenAI is safely and securely embedded into your organization.”
In the case of GenAI and analytics, data governance—paired with a modern architecture—will enable the trust necessary to achieve greater analytics with AI.
Ultimately, enhancing AI automation and processes goes together with easy to find, trustworthy, timely, and protected data. To drive modern analytics, AI and its generative counterpart can play a key role—but only if data governance, security, and compliance is strong alongside it.
Michael O'Donnell (PhD), senior analyst at Quest, revealed the state of most companies operating with data warehouses, lakes, and lakehouses, where these environments are largely used as a single source of truth.
O'Donnell argued that there is room for innovation, explaining that through technological, culture, and process transformation, enterprises can greatly increase data value with a data strategy that aims for self-service, federated governance, data democratization, and organizational agility.
O'Donnell highlighted the following attributes as those that define a modern analytics-focused enterprise:
- Business glossary scaled for governance and compliance
- Access management as the foundation for data democratization
- Data catalog scaled for operationally accurate IT landscape
- Data quality scaled for continuous operating rhythm
- Data modeling that bridges technicalities, enhances clarity, and empowers business insights
- Data literacy scaled through user-friendly asset discovery and a data marketplace
With erwin Data Intelligence by Quest, organizations can adopt this diverse—and foundationally crucial—architecture to forward modern analytics. Combining data cataloging, data quality, and data literacy capabilities to support IT and business needs, Quest’s comprehensive solution delivers enterprise data governance and business enablement at scale.
Dan DeMers, CEO at Cinchy, argued that modern analytics calls for a new approach to data integration. Dubbed “data collaboration,” this strategy aims to deliver co-produced (not shared or integrated) data, which offers real-time collaboration that is productized and federated at scale.
Data collaboration allows data product owners to maintain control while simultaneously acknowledging the complexity of data contribution, even if it crosses domains. The collaborative data products that are generated can act as a “digital twin” for the business, ultimately digitizing how the business works—not simply how the apps work. This drives the “plasticity” of the data, enabling continuous evolution as analytical needs transform without schema lock-in and breaking changes.
For an in-depth review of the changing needs of modern analytics, featuring detailed explanations, demos, and more, you can view an archived version of the webinar here.