Wallaroo Documentation Site

Welcome to the Wallaroo Documentation site! We’ve prepared this site to help you with all of your Wallaroo questions.

If this is your first time encountering Wallaroo, here’s a brief primer. The Wallaroo platform enables companies to manage their Machine Learning (ML) models in a simple, secure, and scalable fashion. The Wallaroo platform offering is comprised of 3 core components:

Wallaroo Components
  • Self-service toolkit for ML model deployment:
    • Integrations: The Wallaroo platform can be installed in any type of environment (cloud, edge, hybrid and on-prem). Additionally, the Wallaroo platform supports ML pipelines across different model training frameworks (TensorFlow, sklearn, PyTorch, XGBoost, etc.). The Wallaroo platform also offers data connectors to process various types of data modalities.
    • ML pipeline management: Data Scientists can leverage the Wallaroo platform’s self-service SDK, UI and API to collaborate, manage and deploy their ML models and pipelines in a production environment.
  • Wallaroo compute engine: Wallaroo’s purpose-built compute engine allows running models on vast amounts of data with optimized computational resource utilization, based on the size of data and complexity of ML pipelines to run.
  • Advanced observability: Data Scientists can generate actionable insights at scale and help identify new business trends by analyzing model performance in real-time within the Wallaroo platform.

For more details on Wallaroo and what it can do for your ML deployments, check out the blog post Meet Wallaroo: The Game-Changing Platform for Deploying Machine Learning

Here’s a quick run down of what you can find in the Wallaroo Documentation site: