Snowflake Introduces Cortex Agents for Managing Enterprise Data
AI agents have the potential to enhance business productivity by automating complex tasks, but their success depends on the availability of high-quality structured and unstructured data. The challenge for many organisations is managing access control, maintaining strict privacy protocols, and efficiently retrieving accurate information.
To address these challenges, Snowflake has launched Cortex Agents, a fully managed service that simplifies the integration, retrieval, and processing of both structured and unstructured data, enabling businesses to build high-quality AI agents at scale.
Mohamed Zouari, General Manager, META at Snowflake, stated: “For AI to deliver real value, enterprises need a robust, governed, and scalable approach to managing structured and unstructured data. With AI expected to contribute up to US$320 billion to the MENA economy by 2030, the region’s rapid adoption of AI-driven solutions demands accurate, secure, and intelligent data management. With Cortex Agents, Snowflake empowers businesses to build high-quality AI agents that drive automation, analytical insights, and company growth. By streamlining data retrieval and processing, enterprises can fully capitalise on AI’s potential while maintaining compliance and being business-ready for the Middle East’s evolving digital economy.”
Cortex Agents, now available in public preview, integrate structured and unstructured data sources, including Snowflake tables and PDF files stored in object storage, to deliver insights. They break down complex queries, retrieve relevant data, and generate precise answers using tools like Cortex Search, Cortex Analyst, and LLMs. This system ensures accuracy, efficiency, and governance at every stage. Cortex Agents can plan tasks, use tools to execute them, and reflect on the results to improve responses, all through a simple REST API for seamless integration into any application.
One key feature is Cortex Analyst, which enables AI-powered SQL generation with semantic understanding. Unlike traditional text-to-SQL systems, Cortex Analyst uses a semantic model to map business terms to underlying data, improving precision in complex, multi-table environments. This tool handles increased schema complexity with advanced JOIN validation, mitigating common issues like hallucinations and double counting. The Analyst Admin UI in Snowsight simplifies semantic model creation, while LLMs help generate semantic models within Snowflake’s secure perimeter, offering customisation options for business-specific needs.
Cortex Search is another core component, retrieving unstructured data such as text, audio, image, and video. This hybrid search engine combines vector and lexical search methods, with semantic reranking, to deliver high-quality, low-latency results. Cortex Search supports indexing millions of rows and has reduced serving costs by 30% due to infrastructure optimisations. It also offers increased customisation, with options to select embedding models and apply advanced filtering capabilities.
AI observability is integral to ensuring the reliability and trustworthiness of generative AI applications. Cortex AI Observability, powered by TruLens, provides end-to-end evaluation of AI agents, assessing metrics like relevance and groundedness. This feature allows businesses to refine agents for improved performance, comparing different configurations to optimise results.
Snowflake’s focus on effective governance and processing of both structured and unstructured data is crucial for creating AI-ready datasets. With its tools for data storage, access, processing, and sharing, Snowflake helps businesses ensure secure, compliant, and efficient AI-driven solutions.
The future of AI agents looks promising as they evolve from simple automation to dynamic systems capable of multi-step reasoning and task execution. By leveraging Cortex Agents, Snowflake provides businesses with the tools they need to harness AI’s potential while maintaining governance and efficiency.
