Author: Paola Di Cretico
-
When “We Can Fix It” Isn’t Enough
NHS data can look like a patient story, but much of it was created for very different purposes. When the data doesn’t behave as expected, fixing it isn’t always the right answer. A mortality data issue made me think differently about data quality, provenance and where the boundaries sit between data management, governance and product…
-
When AI Agents Find Their Own Way Around the Rules
OpenAI’s agents found ways to communicate, reach the internet and repurpose infrastructure that was never intended as an AI tool. The incident raises a bigger question for secure data environments: are we testing what agents are permitted to do, or what they can actually do?
-
A KPI Is Not a Key Result
How do you set meaningful product OKRs when revenue is not the objective and the wider social value may take years to emerge? Drawing on my experience with research data products, I look at how to connect delivery with research outcomes without confusing activity with value or claiming impact the product cannot prove.