Companies utilize Snowflake to store their data in the cloud. With the growing interest in generative AI and large language models, customers seek quick ways to engage with this technology. Today, the company introduced Snowflake Cortex, a fully managed service aimed at facilitating the use of AI-powered applications on the Snowflake platform for both business users and developers.
Depending on your role, Snowflake Cortex serves a couple of purposes. For business analysts, it grants access to various AI tools built on Snowflake’s custom LLMs, simplifying and expediting data interaction within Snowflake. For developers, it assists in creating generative AI applications using the data stored in Snowflake. Partly leveraging capabilities that came to Snowflake through the Streamlit acquisition last year.
At its core, we are incorporating advanced search and large language models directly into the core of Snowflake through a new component we call Snowflake Cortex,” said Sridhar Ramaswamy, SVP of AI at Snowflake, during a recent press roundtable.
“We aim to seamlessly integrate these advanced features, which have become increasingly essential for modern enterprises, deep within Snowflake. This way, our power users, specifically the analysts who extensively work within Snowflake, can significantly enhance their productivity,” added Ramaswamy, who joined the company through the Neeva acquisition earlier this year.
Unlocking the Power of Snowflake Cortex: AI Features and Capabilities
Developers have the opportunity to utilize Cortex for constructing generative AI applications. Simultaneously, the company offers numerous advanced components as ready-made solutions to assist analysts in leveraging generative AI. The first of these is Document AI, A method for extracting data from unstructured documents, such as PDFs and analyst reports. And subsequently querying that data. Sridhar Ramaswamy explained, “Document AI simplifies the process for analysts who may not have specialized programming knowledge or expertise with large language models. It enables them to extract structured data from these documents and organize it into a table.
In practical terms, this allows analysts to query unstructured data contained within these documents.
The second feature being introduced is universal search, a capability that Snowflake acquired when it purchased Neeva in May. Sridhar Ramaswamy noted, “Search, as many people realize, is the foundation for harnessing the potential of language models. We are integrating the core search functionality onto Snowflake objects.” This empowers users to conduct searches across all their Snowflake data and the Snowflake marketplace to discover their data or applications.
The third critical component of Cortex is Snowflake Copilot, which translates plain language inquiries about data in Snowflake into SQL queries. If executed correctly, this has the potential to significantly reduce the time analysts typically spend. Acquainting themselves with data and column structures to construct meaningful queries.
For developers, there are two approaches. They can rapidly build applications using Snowflake models. Alternatively, those seeking greater control over the process can develop customized applications with access to external large language models (LLMs). Such as open-source offerings or those from cloud partners like Amazon Bedrock and Azure OpenAI. Additionally, they can leverage Snowflake Container Services, which were introduced in June, To streamline the deployment of applications as containerized workloads.
Snowflake Cortex plays a pivotal role in the company’s comprehensive strategy to leverage the data. Within Snowflake for various purposes, including search, querying, and application development. Currently, Cortex and its primary features are in a private preview phase, with no specified timeline for a broader release. This is all about Snowflake Introduces a New GenAI Tool That Unifies The Requirements of Developers and Analysts, stay tuned for more info on this post.
CHECK THESE OUT:
- Brave Browser Wants to Help Users Access Tor Easier
- This Data Driven Approach Can Help Your Business Mitigate AI Hallucinations
- Google Marketing Platform – The Benefits and Features of Google Marketing Platform
- Unified Syncing Solution offered by Google Drive for Desktop
- Best Home Décor to Buy This Holiday