Kiana Villaera
Senior AI Architect
KPMG, Microsoft Business Solutions
Data & AI Track
Session Outline:
Machine Learning First, Agents Second
The excitement around AI agents has led many organisations to prioritise generative AI solutions without fully considering their long-term operational costs or business value. As token consumption continues to rise, organisations need to rethink where AI delivers the greatest return on investment.
This session makes the case for putting Machine Learning back at the centre of enterprise AI architectures. Drawing on real-world experience, Kiana explores how predictive models provide explainability, measurable performance, and significantly lower operating costs while solving many of the same business challenges often assigned to large language models.
Rather than presenting Machine Learning and AI Agents as competing approaches, the session demonstrates how they can complement each other. Attendees will see a practical reference architecture where Machine Learning drives the core intelligence and decision making, while AI Agents provide a natural and intuitive user experience.
The presentation concludes with real-world project examples that illustrate how this balanced architecture delivers scalable, cost-effective, and commercially sustainable AI solutions.
About Kiana
Kiana is a Senior AI Architect at KPMG, Microsoft Business Solutions, Malta and a Data Scientist with a Master’s degree from the Asian Institute of Management. She specializes in enterprise-grade predictive analytics, designing data-driven solutions that optimize business KPIs and deliver tangible value.
Kiana approaches every challenge by asking two key questions: “What does the data look like?” and “Will our solution drive profit?”—ensuring a strong alignment between analytical insight and business outcomes.
