I’m excited to share insights from our recent webinar, “Learn Snowflake Cortex AI using Snowpark and Streamlit”.
In this session, I hosted Shankar and Vivekananda, co-authors of “The Ultimate Guide to Snowpark” as speakers. The session explored practical applications of Snowflake Cortex AI and how to use Snowpark and Streamlit to streamline AI-powered data workflows. Here’s a recap of the webinar.
Why Snowflake, Why Snowpark?
Snowflake has come a long way from its data warehousing roots. Today, it stands as a comprehensive AI Data Cloud, a robust solution catering to everything from no-code options to fully customizable AI deployments. This versatility has made Snowflake popular with data engineering, machine learning (ML), and analytics teams, who can tap into its capabilities with ease, scaling efforts quickly and effectively.
Snowpark offers a flexible space to build, train, and deploy AI models, while Streamlit makes it easy to create interactive applications to interpret model outputs. Together, Snowpark and Streamlit help users unlock valuable insights without needing to reinvent the wheel at each stage.
Exploring the Power of Streamlit and Cortex AI
Snowflake Cortex AI is actually a good tool. It simplifies AI operations with tools like Document AI for document processing, Universal Search for data retrieval, and Co-pilot for interacting with the data. Snowflake’s Streamlit allows developers to create applications quickly, with use cases ranging from document processing to managing large language models (LLMs).
Snowflake Cortex is a suite of AI features that use large language models (LLMs). It also includes ML functions that simplify the process of creating and using traditional machine learning models to detect patterns in your structured data. Snowflake ML lets you develop and operationalize custom models to solve your unique data challenges.
Diving into Retrieval Augmented Generation (RAG) with Cortex AI
Vivek discussed Cortex’s Retrieval Augmented Generation (RAG), which makes it easy to enhance AI by giving it access to organizational knowledge. RAG bridges the gap between generalized AI models and organization-specific context, providing more accurate responses based on internal data.
Live Demo of Cortex AI and New App
The demo Vivek shared showcased an open-source toolkit he and Shankar have been developing to make Snowflake Cortex more accessible. Built with Streamlit, this toolkit enables data analysts, business users, and developers to experiment with Snowflake’s AI capabilities without extensive code.
The toolkit (which will be released soon) includes:
- Translation, Summarization, and Text Completion: Users can quickly translate, summarize, and complete text data, all within a user-friendly interface.
- Direct Database Interactions: This functionality allows analysts to access Snowflake data tables directly, making data summarization, completion, and querying seamless.
- And much more
My takeaways
Here’s what stood out most from our session:
- GPU Instances for Intensive AI Tasks: With Snowflake’s GPU-based instances, teams can manage demanding ML tasks efficiently.
- Optimize Data Engineering with Snowpark: By leveraging Snowpark, teams can prepare data faster and more effectively, making it AI-ready.
- Utilize RAG for Contextual Accuracy: RAG is a powerful tool for ensuring LLMs provide contextually accurate responses, especially with organization-specific data.
- Snowflake AI Data Cloud with Cortex AI, Snowpark and Streamlit provides the foundation and recipes for data teams to get started on their own to build apps.
- Vivek and Shankar are Snowpark and Streamlit experts!
Final Thoughts: Simplifying AI with Snowflake
Hosting this session reminded me just how far Snowflake has come in empowering data teams with the right tools for AI-driven projects. Snowpark and Streamlit bring Cortex AI to life, making it easier than ever to work with large datasets, apply contextual insights, and drive real value—all while keeping operational costs in check.
You can view the entire webinar by navigating here.
Thanks to everyone who attended, and to Shankar and Vivek for making complex topics so accessible and sharing such practical insights with the community!
From Revefi’s perspective, Snowflake continues to be a great partner. Our recently introduced, Raden - the world’s first AI Data Engineer helps data teams to optimize data spend, improve data quality, optimize performance for full data observability.
