July 7, 2025

Will AI Solve Universal Data Access

I am hearing the return of the decades-old cry for “universal data access” – particularly in light of the need to train and support AI agents and applications.

While the promise of seamless data flow across diverse systems is enticing, the competitive dynamics of data management providers and the need to rapidly evolve their offerings are inherently in conflict with this vision.

Past initiatives, such as ODBC, JDBC, OLE DB etc etc etc, have undeniably enhanced interoperability, yet true universality remains an elusive vision.

These foundational technologies provided crucial frameworks for building connectors and gateways between disparate applications, tools, data stores, and middleware. They significantly reduced the complexity of integrating diverse data sources.

However, the inherent nature of innovation in the data management sector means that as vendors develop new features, capabilities, and underlying technologies, the “universal” connectors inevitably diverge, requiring constant updates and custom development to maintain compatibility.

Why Do Vendors Resist True Universal Access?

The core reason behind vendors’ resistance to truly universal data access lies in their business models. Data management providers thrive by offering unique, often proprietary, features and performance advantages that differentiate their offerings.

Controlling access to data and applications within their ecosystems creates vendor lock-in, making it more challenging for customers to switch to competing solutions.
This control fosters a competitive advantage, allowing vendors to command higher prices and retain their customer base.

Furthermore, the development of specialized features is often driven by market demand and directly contributes to a vendor’s competitive edge.

Impact On the Market

Instead of universal data access solutions, what we have seen emerge is a significant market of iPaaS solutions, packed with hundreds, even thousands, of pre-built connectors. Focused providers such as SnapLogic, Informatica, Boomi, and CData have emerged to support this need, as well as many of the large data management providers (e.g. Oracle, Microsoft, IBM)

These connectors are the real-world answer to diverse data sources, acting as intelligent translators that bridge the gaps between disparate systems, enabling organizations to achieve the effect of universal data access without waiting for a single, magical standard to emerge.

Will AI Change This Reality?

No and Yes.

I say No because AI and intelligent agents actually depend on the very integration infrastructure that current connectors provide. They also don’t fundamentally change competitive business models of the major data management providers. So they will not solve the challenge of universal data access.

HOWEVER I say yes because what AI and intelligent agents will do is revolutionize how these connections are developed, managed and utilized. We are already seeing this in many of the tPaaS providers (See The Aragon Research Globe™ for the Intelligent tPaaS Market). AI agents are helping to develop connections, they are helping to define efficient indexes and helping developers identify workflows and information (See AI-Enabled Data Lakehouse Architectures).

Bottom Line

The pursuit of universal data access, while laudable, is an unrealistic goal in a rapidly evolving technological landscape driven by competition and innovation.

Instead of chasing this elusive goal, clients must proactively manage and their data architectures so they can accommodate diverse data sources and evolving technologies.

This involves investing in robust integration platforms and implementing data governance frameworks that prioritize accessibility and usability for specific business requirements.

The focus should shift from a utopian “universal” solution to pragmatic, scalable architectures that enable businesses to grow and respond to dynamic market conditions.

UPCOMING WEBINAR

Trends in Corporate Learning: AI Assistants are Here (to Help)

Learning is still a challenge for enterprises. However, the challenge does not end with training employees. In the age of AI, Learning Assistants can help to train people in a variety of ways, and they can also serve as a knowledge base for training AI Agents. In this webinar, Jim Lundy discusses the latest trends in Learning and why the race for outcomes is still the biggest challenge managers face. Key things being covered:

What are the key trends driving learning, and what is the role of the LMS and Learning Content?
What are AI Assistants, and how are they impacting Learning?
What role do AI Coaches play in the race to better outcomes?
How can enterprises gain a competitive advantage by changing how they train?

Will AI Solve Universal Data Access