Wallaroo + Databricks for Rapid Production ML Optimization
Databricks Leaves Teams Wanting More at Production
Data scientists and machine learning engineers like using Databricks for its powerful ETL and end-to-end model training capabilities.
However, they soon find moving those notebooks into production is more manual and compute intensive than anticipated, and they lack the ongoing model observability functions for validating models in production and detecting drift.
Introducing Wallaroo + Databricks
- Deploy, observe, and optimize models in production directly from your Databricks notebook with just a few lines of Python.
- Easily run A/B tests, work together in a collaborative workspace, and monitor drift on all production ML from a single, easy-to-use dashboard.
- All with 10X the speed, and 80% less compute.
Drive Rapid Iteration in Production with the Wallaroo + Databricks Notebook Integration
In this webinar, Wallaroo’s experts walk through our new Databricks notebook integration.
With this integration, data scientists and ML engineers working in Azure Databricks have the easy button for deployment, pipeline management, and tight integration with your Databricks ecosystem for security, authentication, authorization, and data access.
We also walk through a demo showing how to easily take a model trained in Databricks and deploy, manage, and monitor via the Wallaroo SDK.
Jeff Will
Senior Product Manger
Martin Bald
Senior Community Manger