Dive Brief:
- Snowflake launched an agentic AI offering Wednesday as part of a service to simplify data integration, retrieval and processing.
- The vendor’s Cortex Agents can parse requests, analyze unstructured and structured data, ask for clarification and determine next steps within a given application, according to Snowflake. Customers can then track metrics to better understand performance and refine behavior.
- The agents, now available in public preview, are powered by Anthropic’s Claude 3.5 Sonnet model. Use cases include identifying fiscal year start dates, explaining internal naming conventions or prioritizing key tables during SQL generation. Customers can also tailor Cortex Analyst, a SQL translation tool used within the agent, to business needs.
Dive Insight:
Enterprises hope AI agents can improve productivity, but technology leaders are also wary of hype outpacing functionality.
Vendor focus on the latest wave of AI is pronounced, with around 400 providers offering new platforms and related services, according to Forrester market analysis.
Hyperscale cloud providers AWS, Microsoft and Google Cloud added a slew of capabilities to boost adoption in recent months, strengthen governance and ease the development of agents. SAP, Workday, ServiceNow, Slack, Salesforce and now Snowflake are also working to spur agentic AI adoption by lowering the barriers to entry.
While enterprise interest exists, most business decision-makers are still stuck laying the groundwork for adoption. Nearly 9 in 10 IT pros admitted their organization’s tech stack needs upgrading before deploying AI agents, a Tray.ai survey found.
Fortifying governance and security processes are top of mind as well. Gartner predicts AI agent abuse will proliferate over the next three years, underlining the need for investment to protect the business.
Some organizations are already putting AI agents to work. Accenture has around 600 marketing pros leveraging the technology to craft campaigns and assist with other tasks, and it is piloting content production agents. Toyota Motor Corporation is also an early adopter, with around nine agents currently deployed helping to store and share knowledge.
