Find your way to be trained or even get certified in TMAP.
Start typing keywords to search the site. Press enter to submit.
In the GenAI era, Quality Engineering must navigate an explosion of data from both testing AI systems and leveraging AI in testing. The DIKW pyramid offers a structured framework to transform this data into actionable decisions, enabling more effective and intelligent quality engineering practices. The Data, Information, Knowledge, Wisdom (DIKW) hierarchy [Ackhoff 1989] offers a foundational model for transforming raw data into actionable insight. In Amplified Quality Engineering, this model structures how fragmented pieces of data, such as test signals, evolve into informed decisions.
At the foundation is Data: raw, unprocessed facts like log files or test metrics. Data points like a CPU spike or a failed test are meaningless on their own. However, accurate and comprehensive data capture is essential for transparency, traceability, and reproducibility.
Information emerges when data is given context, answering what happened, when, and where. In QA, this might involve linking errors to builds or associating test results with implemented changes. This layer supports operational decisions and traceability.
Knowledge develops when patterns are interpreted. Engineers may discover recurring issues under specific conditions or observe degradation in performance. Analytics and dashboards help reveal these trends.
At the top is Wisdom, applying knowledge with human judgment. It addresses complex questions: Should a release go live if it fails specific tests? How should testing balance risk and business value? Wisdom demands human oversight, ethical reflection, and long-term strategic thinking.
The DIKW model mirrors the evolution of information technology. Early computing emphasized data capture, storing facts without context. As IT matured, it began structuring data into information through relational databases and Management Information Systems (MIS), enabling early forms of evidence-based testing.The knowledge era brought expert systems and BI tools, allowing teams to track anomalies, reuse test assets, and recognize trends. Software QA evolved from record-keeping to strategic insight.
Modern systems now emphasize wisdom through Decision Support Systems (DSS) and ERP, which align decisions with business goals. Quality engineers transitioned from testers to strategic advisors, using data to detect issues and drive what matters most to users and stakeholders.
GenAI reshapes the DIKW hierarchy by accelerating progression through its layers. AI can now parse logs, tag anomalies, and generate summaries, rapidly turning data into information. At the knowledge level, AI discovers trends, such as test flakiness or performance decay, enabling predictive testing and quality forecasting. In GenAI QE, this means detecting behavioral drift or inconsistency across languages and demographics. However, wisdom continues to rely on human judgement. GenAI can simulate scenarios or suggest priorities, but ethical judgment, like weighing fairness against performance, requires human context and accountability.Some propose an intermediate “intelligence” layer, bridging knowledge and wisdom. Here, GenAI shows promise: deriving requirements, assisted coding, composing test cases, analyzing root causes, or adapting to system changes in real time. These capabilities signal a shift toward adaptive, autonomous QA processes that gather information about the system and provide it to the team. The teams use this information to make informed decisions about the quality of the product and the pursued business value.
To succeed in the AI era, quality engineering must evolve to a state where we have experts on quality and testing, and the use of AI:Leverage GenAI to climb the DIKW hierarchy faster, automating data processing, contextualization, and insight extraction.
Maintain human judgment where it matters (the “expert-in-the-lead”), applying experience and ethics to decisions with lasting impact.
Each DIKW layer maps directly onto QA activities:
We see the first steps being taken to apply AI as described above. The expectation is that this will increase in the (near) future, and we will see more and more of these concepts applied in the real world.
AmpQE overview
