Summary
Organizations increasingly shift their focus from merely managing data to creating data products that address real-world problems and deliver measurable value. This change is driven by the recognition that simply accumulating data or building advanced AI capabilities does not inherently yield benefits. Experts emphasize the need for clarity on the specific problems data initiatives aim to solve, as many organizations fall into the trap of prioritizing technology over outcomes.
East Oldfield, a guest on the podcast, highlights the importance of understanding customer needs and iterating on product development through early feedback. He argues that successful data products must not only be valuable and usable but also feasible and viable. The conversation points out that many organizations struggle with defining their data products and often overlook the necessity of engaging with internal users as customers, which can lead to a disconnect between product offerings and actual user needs.
The discussion also touches on the misconceptions surrounding data's intrinsic value, likening it to oil that requires refinement to be useful. Leaders are encouraged to adopt a product mindset, focusing on how data can be transformed into solutions that address specific challenges. As data products evolve, the integration of AI is expected to further differentiate organizations that have solid data foundations from those that do not, emphasizing the need for robust data governance and clear ownership of data assets. Ultimately, the key takeaway is that the value of data lies in its ability to facilitate better decision-making and enhance user experiences, rather than in the data itself.
Show Notes
What turns data into a product people actually want to use? Ust Oldfield joins The Data Hotseat to talk customer thinking, rapid prototyping, and why building better data products takes more than technical expertise.
More from The Data Hotseat
In the hotseat this week is James Bentley, Senior AI Consultant and face of the AI videos on our YouTube channel! With host, Paul McComish, they chat about Genie App Builder - how it differs from other app builders, and how, yet again, the conversation turns to data quality and governance.
In the hotseat this week is Ashley Warren, here to chat about Genie One and Genie Ontology. This episode dives into Databricks' attempt to stop AI from confidently getting business questions wrong. They explore why context is the missing ingredient in enterprise AI, why governance is suddenly everyone's problem again, and whether analysts should be excited, worried, or both.
In the hotseat this week is George Dutton, here to chat about Genie ZeroOps, Databricks' new autonomous background agent for data and AI operations. Is this the beginning of the end for Monday morning firefighting? With host Paul McComish, they dig into how Genie ZeroOps detects, diagnoses and fixes broken pipelines before they hit production, and why sandboxed testing is key to earning engineers' trust.
In the hotseat this week is Peter Sach, who joins our host, Paul McComish, in chatting about Lakehouse//RT and LTAP - two big announcements from Databricks Data + AI Summit. What are they? How do the two integrate with Reyden and the wider Databricks ecosphere? And how could they impact you?
In the hotseat this week is Adam Lammiman, here to chat about the intriguing world of Omnigent, an open source meta orchestration layer for AI agents, recently announced by Databricks. With host, Paul McComish, they cover the rapid evolution of agentic workflows, the hype vs reality of AI advancements, and how Omnigent stands to revolutionise the orchestration of AI processes.