Build better ML models inside Snowflake

Kumo is a Snowpark Container Services App

Solution Brief

Understanding Snowpark Container Services: Enhancing Snowflake's Data Processing Capabilities

Snowpark Container Services (SPCS) is a fully managed container offering in private preview designed to facilitate the deployment, management, and scaling of containerized applications like Kumo within the Snowflake ecosystem. SPCS allows data to be processed without it leaving the Snowflake environment.

snowflake machine learning

About Kumo
inside Snowflake

  • Kumo's intelligent data science solution learns across all of your Snowflake data and eliminates the need for training set generation and feature engineering.
  • From within SPCS, Kumo uses a scalable graph neural network to build ML models and answer business questions.
  • Kumo operates directly on raw Snowflake tables, generates predictions at scale, and writes them back to your Snowflake warehouse.

snowpark container services

Access Kumo through
Snowpark Container Services

Access Kumo inside of Snowflake with a single login to reliably and securely create ML models and make predictions without sensitive data ever leaving your Snowflake warehouse.

data science snowflake

Machine Learning
with Kumo in Snowflake

  • Relational data stored in Snowflake tables hold value that is extremely hard to unlock using traditional methods. To address this challenge, Kumo represents Snowflake tables as a graph and uses graph machine learning directly on raw data tables.
  • Kumo’s graph neural network uses the natural relational structure of Snowflake tables to maximize signal to improve the accuracy and performance of machine learning models and their predictions.

snowflake data science


Deliver More ML Models: Build your graph once by connecting your Snowflake tables, then use it to generate any number of predictions for any number of use cases. Kumo’s automated ML pipelines keep models fresh so the GNN always learns from the latest data.


Improve Performance: Kumo leverages the latest approaches and identifies the best model and parameters for your specific problem and corresponding graph, while still allowing the data scientist to add value at each step. Kumo also improves the accuracy of existing models by directly feeding them trained embeddings.


Accelerate Time to ROI: Because Kumo operates directly on your raw tables, infrastructure costs are reduced. There is no need for ML pipelines, feature engineering, feature stores, and production tooling.


Security and Governance: Kumo inherits all security and governance capabilities from Snowflake through SPCS, and integrates seamlessly with your existing workflows and policies for managing high-value data. Manage your entire ML lifecycle with Kumo inside Snowflake through a single login.

Contact us

to join the private preview for Kumo through Snowpark Container Services.