- Traditional data strategies were 'naive' without crucial context.
- GenAI and 'swarms of agents' are driving unprecedented ROI by taking direct action.
- Google Cloud's optimized infrastructure delivers massive cost and performance gains for AI workloads.
The era of passive data insights is over. Yasmeen Ahmad, Managing Director of Data Cloud at Google Cloud, reveals how the new 'agentic data cloud' is fundamentally shifting organizations from systems of intelligence to dynamic systems of action, powered by generative AI and advanced agent capabilities.
For the past decade, data platforms were primarily built as 'systems of intelligence,' providing dashboards, reports, and predictive scores. However, a significant portion of these insights often remained on the shelf, failing to translate into tangible business action. Yasmeen Ahmad highlights a critical shift: generative AI and agentic systems are now making it much easier to drive action directly into operational and marketing systems, unlocking true ROI.
Ahmad candidly admits that the industry's early approach to AI-ready data was 'quite naive,' focusing predominantly on clean data, lineage, and quality. While important, this only achieved 50% accuracy with agents. The missing half, she explains, is context – the 'invisible work' of business intuition and hidden meaning that was previously confined to human minds. Google Cloud's knowledge catalog now infers this schema and meaning across structured and unstructured data, enabling agents to reason over data accurately and efficiently. Gemini Enterprise acts as the new 'front door,' allowing business users to converse with their data without grappling with underlying complexities, further enhanced by deep research agents that can reason over both enterprise and web data for holistic insights.
The scale of this agentic transformation demands robust infrastructure. Ahmad notes that agents can generate 10 to 20 API calls for every human click, necessitating a focus on performance and cost. Google Cloud is addressing this with significant innovations, including a 35% increase in BigQuery query processing speeds and a 40% reduction in cost over the last year. Furthermore, the managed service for Apache Spark with the Lightning engine is five times faster than vanilla Spark. The company's vertical integration, from infrastructure to models and data, allows for unparalleled stack optimization, exemplified by a 230x reduction in token usage for AI inferencing over BigQuery data.
Addressing the multi-cloud reality, Google Cloud's cross-cloud lakehouse leverages open standards like Apache Iceberg and cross-cloud interconnect technology. This allows customers to connect and universally view data regardless of where it resides – be it AWS, Azure, Databricks, or Snowflake – without costly data movement or proprietary formats. This open approach, combined with the Data Agent Kit (providing plugins, extensions, and tools), empowers agents to natively understand and optimize Google's data cloud, ultimately driving the shift from mere data understanding to impactful, real-time action.
“I'm super excited to see the system of of intelligence move to systems of action. And in particular here at next, I've now heard three, four different use cases from customers just this morning of how they are engaging swarms of agents that are driving true action.”
- Yasmeen Ahmad, Managing Director, Data Cloud
