Skip to content Professional EditionNEW
Purpose-built environment for small data teams and dbt Core developers. Enterprise Edition is the leading provider of Snowflake environment management, end-to-end orchestration, CI/CD, automated testing & observability, and code management, wrapped in an elegant developer interface.
Spendview for Snowflake FREE

An inexpensive, quick and easy way to build beautiful responsive website pages without coding knowledge.

Pricing and Edition

See whats included in our Professional and Enterprise Editions.

Getting Started
Docs- New to DataOps.liveStart learning by doing. Create your first project and set up your DataOps execution environment.
Join the Community
Join the CommunityFind answers to your DataOps questions, collaborate with your peers, share your knowledge!
#TrueDataOps Podcast
#TrueDataOps PodcastWelcome to the #TrueDataOps podcast with your host Kent Graziano, The Data Warrior!
Resource Hub
On-demand resources: eBooks, white papers, videos, webinars.

Customer Stories

Enroll in the Academy to take advantage of training courses. These courses will help you make the most out of

Learning Resources
A collection of resources to support your learning journey.
Connect with fellow professionals, expand your network, and gain knowledge from our esteemed product and industry experts.

Stay updated with the latest insights and news from our DataOps team and community.
#TrueDataOps is defined by seven key characteristics or pillars:
In The News

In The News

Stay up-to-date with the latest developments, press releases, and news.
About Us
About UsFounded in 2020 with a vision to enhance customer insights and value, our company has since developed technologies focused on DataOps.


Join the team today! We're looking for colleagues on our Sales, Marketing, Engineering, Product, and Support teams.
Thomas Steinborn, SVP ProductsMay 13, 2024 12:21:18 PM3 min read

Building Data Applications using Snowflake Cortex ML and LLMs

Building Data Applications using Snowflake Cortex ML and LLMs

Overview of Snowflake AI/ML 

Snowflake provides a rich set of Artificial Intelligence (AI) and Machine Learning (ML) capabilities covering various use cases. 

Snowflake AI, Snowflake ML, and Snowflake LLM summary 

Cortex LLM 

Cortex ML 

Cortex Search Service 

Document AI 

Snowflake Copilot 

Snowpark ML 

Snowpark ML Packages 

Snowpark Model Registry 

Snowpark Feature Store 

Snowpark Container Services (with Nvidia GPUs) 

Snowpark provides great flexibility in your choice of data science tasks. You can choose which Python packages to use, which models to run, and, of course, use the ever-popular Pandas. 

Snowpark Container Services (SPCS) gives more freedom and is highly effective for AI tasks when used with GPU compute pools. One common scenario is to train a model on SPCS and then use Snowpark ML to run predictions. This offers a good balance between cost and performance. 

Snowflake Cortex functions provide serverless ML and LLM functions on top of your data in Snowflake. Let’s focus on Cortex and how we help you to build data applications rapidly. 

What can you do with Cortex LLM functions and DataOps? 

Snowflake Cortex LLM functions cover common use cases for text analytics and chatbots. 

Let's start with text analytics against transcription of all your meeting recordings. Snowflake simplifies access to accurate summaries without any prompt engineering. A simple call to SNOWFLAKE.CORTEX.SUMMARIZE against your table of transcription is sufficient. 

Within, you can develop a full Streamlit application calling the native function from Python: 

 0 - Streamlit Python code - zoom

The final result can be a rich user experience built with and deployed as Streamlit in Snowflake. Based on the summary, you can then choose a meeting recording and analyze it: 

 1 - Streamlit-video-analyzer-summarize

Once you find a recording that interests you, start interacting with it. 

Let’s create a chatbot to query the full transcription in natural language. The SNOWFLAKE.CORTEX.COMPLETE function is the right choice to pass your input as prompts to a Large Language Model (LLM). 

 2 - Streamlit chat code - zoom

Snowflake offers the choice of different LLM models to you to tailor it for your use case 

  • mistral-large 
  • mixtral-8x7b 
  • mistral-7b 
  • llama2-70b-chat 
  • gemma-7b 

For our example, mistral-large gave the best results. 

In addition, you can use the just-announced Snowflake Arctic model. 

Build and deploy it with and provide a fully immersive experience to your users: 

 3 - Streamlit-cortex-chat


What can you do with Snowflake Cortex ML functions and DataOps? 

Cortex further provides ML-based functions working on top of your Snowflake data. You can benefit from further use cases, e.g., time-series forecasting or anomaly detection. Time-series forecast employs a machine learning algorithm to predict future data using historical time series data. Anomaly detection is the process of identifying outliers in data. 

When you want to use time-series forecasts, you can use the Snowflake Object Lifecycle Engine (SOLE) to create your data tables, run your data pipeline to ingest the necessary data and then launch into our development environment Develop. 

You can explore the underlying Snowflake data directly in our browser-based IDE. We will use a Jupyter Notebook to connect to Snowflake. Then, we will run a Pandas query with Snowpark on your data table. 

10 - explore the data with Develop

Once you reviewed the data, plotted it, and found the interesting data pattern, you can prototype the forecast and visualize the upper and lower bound as well as the expected forecast for the next few months. 

11 - forecasting a time series with Develop

Next, you can create your Snowflake view for your training data. Once done, you can create your SNOWFLAKE.ML.FORECAST function my_forecast_model. Later, you can use the new function in standard SQL with CALL my_forecast_model!forecast. 

Learn more about the platform and its capabilities here. 


Thomas Steinborn, SVP Products

Thomas comes to with over 20 years of experience in integration technology ranging from application integration over data integration to data governance. He believes in DevSecOps principles as applied to Data Products and Data Operations (DataOps) and has brought them to the products he’s led since 2005. Thomas has spent over ten years scaling product management organizations at the global level. Before DataOps, he was Chief Product Officer at Infoniqa, the Human Capital Management, Payroll, and Time Solutions leader in the DACH region. Before that, he led product and user experience as VP of Products at Talend, a leading provider of data integration, data integrity, and data governance solutions.