Blog

Catalog: The launchpad for unstructured data management

Labelbox built Catalog, a unifying layer across your ML data, to tackle managing massive unstructured datasets, visualizing and exploring data, and curating efficient batches of high-value data.

6 fundamental sessions on improving your AI data

Join us at Labelbox Blueprint on May 18th - A virtual event covering six fundamental sessions to improve your AI data

Streamlined data management and key editor improvements

We've focused recent product development on streamlining essential data management workflows and are introducing new collaboration tools for labeling operations and labeling teams. We’ve also made some key updates to our Editors that enable more complex ML use cases and improve usability.

March updates

We’ve focused recent product development on improvements to Catalog, the ability to filter data rows by status, updates to our account usage page, and more.

How to use vector embeddings to create high-quality training data

Functions and embeddings are a powerful way to quickly search and explore your unlabeled and labeled data. Using them can help you break down silos across datasets so you can focus on curating and labeling the data that will most significantly improve model performance.

How to improve model performance with less data

This data-centric ML method can reduce the required amount of training data by 10% to 50%.

Stop labeling data blindly

ML teams need to know "what data will improve model performance fastest?" These best practices will help teams identify model errors and prioritize better data.

You’re probably doing data discovery wrong

Learn best practices for storing, evaluating, and visualizing labeled and unlabeled data so you can accelerate your model iteration cycles and simplify your training data pipeline.