Building Generative AI Apps to Talk to Your Data
2h 17mIntermediate2025-05-29
Authors
Snowflake, Inc
Course details
Explore the power of generative AI in transforming how you interact with enterprise data by building conversational applications. Discover how to use retrieval-augmented generation (RAG) for unstructured data and table-assisted generation (TAG) for structured data to create intelligent, dynamic interfaces. Learn to leverage Snowflake's Cortex platform to develop applications that answer user queries in natural language, crossing the boundaries of traditional data analysis. Understand how to create semantic models that improve SQL generation; dive into practical implementations with Cortex Search and Cortex Analyst; and learn how to enhance your workflows with efficient data querying and processing techniques. Whether you're a data scientist, AI engineer, or business analyst, this course offers hands-on experience using Snowflake's built-in tools to optimize data interactions and unlock hidden data value.
Skills covered
SnowflakeTelecommunicationsCloud StorageNatural Language Processing (NLP)Full-Stack Web DevelopmentDatabase DevelopmentDatabase ManagementCloud ServicesArtificial Intelligence (AI)Web DevelopmentNetwork and System AdministrationCloud ComputingSoftware DevelopmentOne-Off
Concepts
0. Introduction
- 01 - Getting answers from data
- 02 - Getting answers from data - Using RAG
- 03 - Getting answers from data - Using text-to-SQL
- 04 - What we will build - Success in this course
1. Introduction for Conversational Apps with Data
- 05 - Querying unstructured data - RAG
- 06 - Querying unstructured data - Hybrid search
- 07 - Querying structured data - Text-to-SQL
- 08 - Let s dive in
2. Finding Answers from Unstructured Data with Cortex Search
- 09 - Opening up unstructured data with RAG
- 10 - What is Cortex Search
- 11 - Loading unstructured data to a Snowflake stage
- 12 - Parsing and chunking text
- 13 - Creating the Cortex Search Service
- 14 - Building a RAG with Cortex Search
- 15 - How to measure success
- 16 - Automatic processing of new documents
- 17 - Building out the front end
- 18 - Well done
3. Answer Questions from Structured Data with Cortex Analyst
- 19 - Avoiding death by dashboard
- 20 - Setting up the app
- 21 - Understanding the semantic model
- 22 - Working with the semantic model
- 23 - Using Cortex Analyst
- 24 - From text-to-SQL to TAG - Creating table-assisted generation
- 25 - Expanding the scope of the semantic model
- 26 - Building the Streamlit app
- 27 - Congratulations