Post Detail Image
Home
Blog
January 2026 Feature Updates: Smarter Reviews, Better Search, and Expanded Machine Learning Support
Contents
Datasaur

January 2026 Feature Updates: Smarter Reviews, Better Search, and Expanded Machine Learning Support

January’s release: data review faster and more transparent with smarter search, clearer review signals, and better control over pre-labeled data. We’ve also expanded ML-assisted labeling with Amazon Bedrock support for greater flexibility.
by
Datasaur
on
February 3, 2026

January Feature Updates: Smarter Reviews, Better Search, and Expanded Machine Learning Support

Reviewing data shouldn’t feel like detective work. This release brings clearer signals, smarter search, and more control over machine-generated labels so you can spot issues faster and focus on what actually needs review.

More Control Over Pre-Labeled Data: Review and Approve Labels in Bulk

When “Require approval for pre-labeled labels” is enabled, users can now view a dedicated list of pre-labeled labels and accept or reject them individually or in bulk via the Labels extension, keeping quality checks fast and explicit even when working with large volumes of pre-labeled data.

Learn more

Smarter Search in Span Labeling: Filter by Label Status

The Search Extension in Span Labeling now includes additional filtering options, allowing users to refine search results by label status and quickly narrow in on labels that need review or verification.

Learn more

More Flexibility in ML–assisted Labeling: Amazon Bedrock Support

ML–assisted labeling now supports Amazon Bedrock as an LLM provider, giving teams more flexibility to choose models that align with their infrastructure, compliance, or cost requirements.

Faster Navigation: Search Your Joined Workspaces

The joined workspace list now includes a search function, making it easier to find and switch between workspaces, especially for users who collaborate across multiple teams or projects.

Clearer Reviews: Visible Answer Count for Multiple-Answer Questions

Multiple-answer questions now display a visible answer count, making it easy to see how many answers exist at a glance.

Refined Span Labeling Experience: Clearer Selection and Improved Readability

Span Labeling UI improvements include clearer visual feedback when right-clicking to select a line and updated label information tooltip colors, making selections more obvious and label details easier to read during labeling or review sessions.

Most of this release is about clarity: clearer reviews and more precise search. Alongside that, Amazon Bedrock joins as a new Machine Learning provider for ML-assisted labeling.

No items found.
Related post
Post Image
Datasaur's 2025 Feature Releases: Powering Smarter, Faster Data Labeling
February 3, 2026

As we move past 2025, the Datasaur team has been hard at work delivering features that make data labeling more efficient, flexible, and intelligent. Whether you're annotating text for NLP models, managing complex labeling workflows, or building datasets for LLM training, this year's updates are designed to help you work smarter. Here's a rundown of what's new and why it matters.

Datasaur
Post Image
December Feature Updates: More Flexible Rows, Hybrid Labeling, and Expanded Evaluation Metrics
January 5, 2026

This release brings more control to structured data workflows, new flexibility in how you label data, and broader support for evaluating span projects. Here’s what’s new.

Datasaur
Post Image
November Feature Updates: Cleaner Span Editor, Better Search, and More Accurate Time Tracking
December 7, 2025

This month’s release focuses on a tidier labeling workspace, tighter review controls, and more accurate time reporting, especially when AI assists the work.

Datasaur
See more on the blog