Exciting new tools

1. Cortex AISQL: Seamlessly Integrating AI into SQL

At the conference,  Snowflake dropped something that’s going to change how analysts and data teams work with AI: Cortex AISQL. Here’s what you need to know.

Powerful AI Functions in One Line of SQL

Cortex AISQL brings a set of native SQL functions like AI_COMPLETE, AI_FILTER, and AI_CLASSIFY, making it easy to run tasks like summarization, classification, or semantic search with a single query. Thanks to support for a new FILE datatype, these functions work not just on text, but also on images and documents.

You can insert an image or audio file directly into your SQL query, and use it as a filter. For example, finding similar images or matching voice recordings to known patterns. This enables use cases like product image tagging, contract summarization, and natural-language filtering of user feedback, all from within Snowflake.

Fast, Cost-Effective, and Built for Unstructured Data

According to Snowflake, Cortex AISQL delivers faster performance and lower costs, up to 70% faster and 60% cheaper than traditional AI workflows. It’s also a major simplification for working with unstructured data. Instead of building complex pipelines to extract meaning from raw text, documents, or images, you can now handle it directly in SQL. Whether you’re tagging sentiment in reviews or pulling structured fields from PDFs, Cortex functions like AI_CLASSIFY and AI_SUMMARIZE_AGG turn messy data into usable insights with a single query. That means faster development, lower costs, and fewer moving parts in your data stack.

2. Snowflake Intelligence: Natural Language Meets Your Data Warehouse

Snowflake are also changing how people interact with data by introducing Snowflake Intelligence, a conversational AI interface that lets anyone query their data using plain language:

Natural Language Meets Real-Time Data

With Snowflake Intelligence, users can ask plain-language questions like “What were our top-selling products last month?” and get instant answers from live Snowflake data. It automatically writes and runs the right queries using semantic models, which define metrics and dimensions in terms your organization already uses. These models also support synonyms, so different teams can refer to the same metric using their own language (like “sales,” “revenue,” or “top-line”) and still get consistent results. It’s a flexible, human-friendly way to work with data across departments.

Built for Security, Governed by Design

Unlike generic AI chat tools, Snowflake Intelligence is built on top of your existing data governance. It enforces role-based access, honors data masking rules, and never exposes unauthorized information. Powered by trusted LLMs like OpenAI and Anthropic, it works seamlessly with Snowflake’s new Semantic Views to deliver business-aware, secure insights in seconds.