Tech & AI LIVE: Gen AI – Building Robust AI, Kevin Fleet

December 09, 2024

undefined mins

Kevin Fleet, Vice President of Technical Sales Strategy and Operations at Informatica, explores building robust AI with data at Tech & AI LIVE: Gen AI 2024

Kevin Fleet, Vice President of Technical Sales Strategy and Operations at Informatica, captivated the audience at Tech & AI LIVE: Gen AI with his keynote on creating relevant, responsible and robust AI through better data management.

Sharing insights from over 25 years in the data management industry, Kevin highlighted the critical role of ‘AI-ready data’ in driving successful AI initiatives.

The importance of AI-ready data

Kevin opened with a thought-provoking question: What happens when bad data fuels AI systems?

Using relatable examples, such as vacation planning gone awry, he illustrates the significant consequences of flawed data. In business contexts, these missteps can be catastrophic.

Citing Zillow’s US$300m miscalculation in home resale values—which led to US$1 billion in market capitalisation loss—he emphasises, “If it can happen to Zillow, it can happen to anyone.”

Despite advancements in AI models, Kevin notes that organisations often underinvest in data preparation, spending disproportionate time on model development rather than ensuring data quality.

“Bad data leads to bad AI,” he warns, underscoring the need for robust data management as a foundation for AI success.

Kevin Fleet, Vice President of Technical Sales Strategy and Operations at Informatica

Characteristics of AI-ready data

For AI systems to deliver on their promise, Kevin outlines three essential qualities of AI-ready data:

Relevant: Accurate, contextual, and transparent data powered by universal metadata.
Responsible: Governed, secure, and unbiased data that aligns with compliance and privacy standards.
Robust: Scalable, complete, and resilient data that supports enterprise-scale AI applications.

Achieving these qualities requires a concerted effort in data profiling, stewardship, governance, and ongoing quality management.

Kevin Fleet, Vice President of Technical Sales Strategy and Operations at Informatica

The role of Informatica in AI success

Kevin describes Informatica’s Intelligent Data Management Cloud as a powerful enabler for AI-ready data. Powered by the AI agent ‘Clair’, the platform integrates advanced data management capabilities, including data cataloguing, governance, and lineage tracking.

By analysing over 90 trillion monthly transactions, Clair provides actionable insights to enhance data quality and drive AI initiatives.

Informatica’s latest innovation—a GPT-powered interface—simplifies data access and utilisation.

“It’s a single conversational interface that connects to your enterprise data, enabling users to ask complex questions and generate actionable insights without writing code,” Kevin explains.

Kevin Fleet, Vice President of Technical Sales Strategy and Operations at Informatica

Overcoming the fear of missing out

Acknowledging the rapid pace of AI innovation, Kevin urges organisations to avoid rushing into generative AI transformations without addressing foundational data issues.

“This pressure to move quickly opens up organisations to significant risks,” he says.

Kevin stresses the importance of making strategic investments in data infrastructure, governance, and quality to ensure sustainable AI success.

Data leaders, he argues, have a unique opportunity to advocate for these investments, leveraging high-profile AI projects to highlight the critical role of data management.

“If there’s one AI project today, there will be ten next year and thirty the year after,” he predicts, emphasising the need for scalable, repeatable systems.

Kevin’s keynote reinforces a crucial message: the success of AI depends not just on cutting-edge models but on the quality of the data that fuels them. By prioritising relevant, responsible and robust data practices, organisations can unlock the full potential of AI while mitigating risks.

As he concludes, “Bad data leads to bad AI.” For organisations aiming to thrive in the AI-driven future, the path starts with getting their data house in order.