Viewpoint: Foundations for Generative and Agentic AI
March 16, 2026
Since the introduction of generative AI, many organizations moved quickly with writing assistants, code copilots, and internal chatbots.
While experimentation is widespread, scaled and measurable value is not universal. Organizations that rushed into generative AI are proceeding more cautiously with agentic AI. The move to systems that reason, invoke tools, and take action raises the stakes.
The roadblock for multi-agent systems is not model access; it’s the data foundation. Siloed systems, inconsistent metadata, brittle pipelines, unstructured data quality, and lack of unified platforms remain obstacles.
To explore how vendors are responding to these realities, TDWI spoke with Prasanna Krishnan, who leads several product areas at Snowflake including Marketplace, data and AI sharing, Postgres, and Horizon, the company’s governance and catalog platform.
Download this special Viewpoint today to learn how you can create the right supports for sustainable generative and agentic AI deployments.