Snowflake Agentic AI: Cortex Analyst, Search & AI Agents

Create a Snowflake account from start to finish, enabling a $400 free credit, choosing the standard edition on AWS, completing registration, activation, and login, with Apache Iceberg ready.

Explore Snowflake's cloud native data warehouse architecture, featuring separation of compute and storage, elastic multi-cloud scalability, and secure, concurrent analytics across AWS, Azure, and Google Cloud.

Create Snowflake database tables and establish your data connections via sql worksheet. Explore the Snowflake sample data and tpc-h sf ten schema with customer and orders to power ai agents.

Copy seven tables from the Tpc-h SF ten schema into the sales public schema, using create table as select, enabling isolated analytical environments, batch processing, and data verification.

Discover how to create and manage a Snowflake stage with SQL, enabling storage for semantic models and YAML configurations, and use the stage dashboard to enable directory table.

Explore Snowflake's AI and ML menu, from Cortex Playground to Cortex Analyst and Cortex Search, and see how intelligent agents orchestrate data insights in a conversational, automated workflow.

Enable regional AI models in Snowflake by flipping the switch to Cortex cross-region inference, access models like Claude and ChatGPT, and verify in Cortex Playground for advanced analytics.

Explore AI agents and autonomous intelligence systems, detailing perception, reasoning, action execution, planning, and learning to enable independent, goal-driven decision making and real world impact.

Snowflake agents integrate cortex search and cortex analyst into a unified framework, orchestrating structured and unstructured data for conversational enterprise analytics. They handle csv, pdf, and time series data.

Explore how Snowflake Analyst enables natural language conversations with enterprise data, delivering real-time analytics, semantic understanding, and intelligent decision support for conversational analytics and business intelligence.

Create a single table semantic model for customer analytics, selecting sales data and all columns to enable deep conversational insights with transparent sql and metadata.

Learn how to build a multi-table semantic model for customer and order data, enabling cross-table insights and natural language queries with automated metadata and a yaml footprint.

Explore semantic views for single-table analytics to answer targeted business questions with rapid deployment, then create a customer view by selecting database, table, and columns for natural language analysis.

Explore ten components of Snowflake Cortex Agents architecture, from the intelligence layer to semantic processing, data integration, tool orchestration, security, governance, and deployment for conversational analytics and enterprise analytics.

Create a Cortex agent in the Snowflake Intelligence Database with a single tool. Configure the Customer Analyst using the Semantic Model and the Customer Performance YAML on My Stage.

Explore Snowflake intelligence interface and its chat-like data analytics experience, including accessing the customer agent, querying in plain English, and discovering table columns and data types.

Transform the cortex agent into a multi-table analytical assistant by adding orders data and a second analyst to enable cross table insights.

Explore analyst source management to dynamically switch between the customer analyst and customer agent analyst, choosing sources with distinct data scopes to optimize cross-table analysis and performance.

Select and configure the orchestration model to balance performance, regional availability, and analytical requirements, choosing automatic or Claude or GPT variants for optimal conversational analytics, with governance and cost considerations.

Learn to customize agent communication through response instructions, defining formats and behavior in the agent configuration interface. Configure outputs as json, csv, tabular, or graphical visuals for seamless business integration.

Configure default questions to guide users with pre-configured analytical prompts, enabling immediate conversational analytics, reduced cognitive load, and hybrid access to both sample and custom queries.

Build a comprehensive all-tables semantic model in Snowflake Cortex Analyst, enabling cross-table, natural language business intelligence and an intelligent agent for enterprise-scale conversational analytics.

Create an enterprise sales agent that analyzes customers, orders, products, suppliers, nations, and regions through a comprehensive semantic model and natural language queries for cross-table business insights.

Explore how Snowflake Intelligence converts natural language questions into interactive line, bar, and table visualizations using agent-based conversational analytics.

Explore how Snowflake Cortex Search enables intelligent, conversational discovery across enterprise data with natural language processing, semantic understanding, real-time results, and secure, multi-source access.

Explore traditional keyword search without ai, from user queries and tokenization to inverted indexes and tf-idf ranking. Understand how precise term matching, boolean logic, and document delivery drive reliable results.

Explore rag retrieval and augmented generation, grounding responses in organizational knowledge via vector databases and embeddings, as the front end passes through api gateway to a large language model answer.

Discover how vector embeddings transform text into dimensional vectors across six layers, including data input processing, semantic encoding, and similarity metrics, to enable semantic search, content discovery, and business intelligence.

Compare cortex search with traditional search, highlighting AI-powered natural language processing, semantic understanding, unified data integration, and conversational interfaces for scalable, context-aware discovery.

Explore cortex search architecture, merging AI-powered retrieval and generation with document ingestion, embeddings, and vector search on snowflake. Learn how query APIs coordinate ingestion and semantic similarity to deliver answers.

Learn to upload pdf documents to Snowflake stages, create a dedicated stage, and enable text extraction, content analysis, and intelligent search within the Snowflake data platform.

Learn how to extract embeddings and metadata from PDF documents using Snowflake cortex, converting unstructured PDFs into structured tables with OCR, content chunking, embeddings generation, and a document search service.

Learn to perform direct document search with SQL on Snowflake Cortex, enabling programmatic integration with apps, workflows, and analytics using parse_json and search_preview.

Build an AI powered document search service in Snowflake using Cortex Search. Configure from a PDF data source and use version two embedding for semantic, natural language search.

Enhance the sales agent within Snowflake Cortex Analyst with a unified search service, enabling seamless, conversation-based analysis of both structured data and unstructured documents.

Discover Snowflake cortex AI and LM functions, enabling SQL-based access to GPT and Llama for text analysis, generation, translation, extraction, and classification inside Snowflake's secure environment.

Learn to use Snowflake Cortex summarize to condense earnings reports and other texts into precise, key-number summaries, enabling scalable bi workflows and sql-based analytics.

Explore sentiment analysis with Snowflake Cortex sentiment function, from basic scores to advanced aspect-based and multilingual analyses across regions. Integrate these capabilities into your business intelligence workflows for maximum impact.

Explore the complete function in snowflake cortex and Arctic model for text generation. Master prompts, batching, and settings like temperature and max tokens for onboarding and educational content.