40% of enterprises will scrap AI agents - 3 ways to ensure yours don't fail (www.zdnet.com)

At the Snowflake Summit in San Francisco, digital leaders from diverse organizations shared insights on deploying AI agents effectively, emphasizing the importance of structured frameworks, expert involvement, and data monetization. Matt Luizzi from Whoop discussed how they've scaled their use of the Snowflake CoCo coding agent through formalized evaluation frameworks, which have allowed them to rapidly iterate and enhance their service delivery. The organization's centralized data infrastructure facilitated this process, underscoring the need for strong contextual frameworks in AI implementation. Madeleine Want of Fanatics highlighted the critical role that high-quality data and governance play in the efficacy of AI agents, revealing a successful shift from bespoke models to integrating third-party systems. This paradigm shift has fostered broader applications of AI, with users now demanding more operational use cases. Sriram Sitaraman from Synopsys reflected on the tangible benefits of AI agents, noting that they have improved decision-making quality while optimizing time and cost metrics. Collectively, these experiences reveal a clear path for enterprises to capitalize on AI technologies, ensuring that such initiatives are built on solid data foundations and expert knowledge.