AWS Bedrock User Level Attribution

CloudHealth now supports allocation of Bedrock costs by IAM Principle and Inference Endpoint. These new dimensions have been added as categorizations on the AWS All AI dataset. Two new columns, Principal Type and Principal have been added to capture Users, Assumed roles, IAM principals and inference endpoints.

To enable these columns, customers must be using CUR 2.0 and have enabled the IAM Principal column in the CUR 2.0 Configuration. Datasets will backfill for all periods where CUR 2.0 has been collected. 

Note that enabling IAM Principals in the CUR will bring in user names and email addresses to CloudHealth. At this time there is no way to disable user names and emails coming into CloudHealth from the CUR if you enable this setting. If you do not wish to have email addresses displayed in the platform, do not enable this setting. 

Data Connect - Support for GitHub Copilot 

GitHub Copilot Enterprise is now an available datasource in Data Connect. Customers can now report on the cost and usage per team, project, user and organization. 

The GitHub Copilot Cost and Usage dataset will be available in the datasets tab and will back the  standard report. 

This can be configured by navigating to Data Connect → New Datasource → Github Copilot. To configure, an Account ID and PAT token are required. 

Data Connect - Support for OpenAI Enterprise

The existing OpenAI connector now supports an additional configuration option - OpenAI Enterprise. When configuring the OpenAI connector, select the appropriate agreement type from the dropdown before continuing configuration. 

Both OpenAI agreement types will write to the same OpenAI cost and usage dataset. OpenAI Enterprise gives us two files, with the same cost and usage, pivoted by Client and Model. When viewing the combined OpenAI Cost And Usage Report, which combines Platform and Enterprise Agreements into one view, Model Cost and Client Cost have their own measures so as to not double count. To view cost or token consumption by Client, select the correct measure prefixed by “Client”.

OpenAI Enterprise will also generate two standard reports and datasets, OpenAI Enterprise Cost by Model and OpenAI Enterprise Cost by Client

Dashboard Improvements

To support better dashboard performance, several changes have been made. 

After today, out of the box dashboards, known as Standard Dashboards, may no longer be customized. Customers must clone standard dashboards in order to make changes. Any standard dashboards that have already been customized have automatically been cloned and appended with “-copy”. No work will be lost and default dashboard settings will be persisted. The only dashboard actions are now “Clone” and “Make Default”.

When editing a dashboard and accessing the widget library, any widgets that enable custom content authoring will now be pinned to the top of the library. Including KPI, HTML, Custom Report and Section. 

Dashboards now support a new widget type called “Tab”. Customers can add multiple tabs to dashboards, and widgets can be placed in each tab. Tabs function similar to Sections, and can contain any number of widgets. 

Credential Management for Data Connects

Previously only available through API, CloudHealth has added Credentials to the new experience. Credentials were introduced to simplify managing multiple Data Connects collecting from the same location. These can now be managed directly when configuring a Data Connect or from the separate Credentials menu. Fore more details on Credentials, read more here

GreenOps data powered by Greenpixie 

CloudHealth has announced a strategic partnership with an industry leader in GreenOps data, Greenpixie. With the new integration, Greenpixie now supplies the emissions data and methodology behind CloudHealth’s GreenOps dashboard and emissions dataset. With this change, CloudHealth has improved it’s accuracy in reporting on carbon emissions and power consumption verified to ISO 14064 and now includes Scope 3.

No action is required, this data began flowing automatically. For more details, check out our documentation here

Deletion of Data Connect Connectors

You can now delete unwanted, test, or misconfigured Data Connect integrations directly from the platform. When deleting a Data Connect, you have the option to permanently remove the connection along with all its downstream dependencies, including ingested datasets, linked reports, and historical configurations, preventing orphaned resources and storage clutter. This deletion workflow is also the supported method for updating custom dataset schemas, which cannot be edited after initial creation.

Note: Deleting a Data Connect and its associated resources is permanent and irreversible. Confirm that all historical data and dependent reports are no longer required before proceeding.

FlexOrg Support for Data Connect

Customers using FlexOrgs can now segment access to Data Connect data the same way they segment cloud accounts. All Data Connect types (including Custom File Uploads) can now be assigned to FlexOrgs and costs will be filtered by organization.

For custom datasets, a FlexOrg Mapping Column can be selected when configuring the Data Connect. Each value in that column can then be assigned to a FlexOrg. This lets you segment custom data by any dimension in your file. If that dimension in the custom file maps with either an AWS Account, Azure Subscription, or GCP Project, you can also take advantage of FlexOrg assignments previously done for those clouds.

To assign Data Connect accounts for newly supported providers, navigate to the FlexOrgs screen in the Classic experience. You’ll find all your Data Connect providers alongside your existing cloud accounts.

To designate a column to use for FlexOrg segmentation for a Custom File Upload, edit an existing Data Connect schema within the new CloudHealth experience.

Explore Datasets 

Exploring Datasets is now easier. Creating a report directly from Datasets with the “Create Report” Option on the top right end of the Dataset Sample View. Reports created using this option allow users to switch granularity and explore all Datasets. Read more about it here 

AWS Standard Reports added to the new CloudHealth Experience (EBS, S3, DynamoDB, EBS, Redshift)

The New experience now supports Standard reports and Datasets for the following: 

• AWS S3 Cost: Provides detailed effective cost data for AWS S3 by resource type at the Hour, Day, Week, and Month granularities.

• AWS EBS Volume Cost and Usage: Provides detailed effective cost and usage data for AWS EBS by volume type at the Hour, Day, Week, and Month granularities.

• AWS DynamoDB Cost and Usage: Provides detailed effective cost and usage data for AWS DynamoDB by billing account name at the Hour, Day, Week, and Month granularities.

• AWS Redshift Cost and Usage: Provides detailed effective cost and usage data for AWS Redshift by reservation type at the Hour, Day, Week, and Month granularities.

• AWS ElastiCache Cost and Usage: Provides detailed effective cost and usage data for AWS ElastiCache by reservation type at the Hour, Day, Week, and Month granularities.

FOCUS columns like Effective cost, Billed Cost and Region Name continue to be available as common measures and dimensions across all Standard reports. 

AWS CUR 2.0 Support

CloudHealth now supports collecting and reporting off of AWS CUR 2.0 using AWS Data Connect. With this release, CloudHealth also exposes the IAM Principal column associated with Bedrock User Attribution and makes this available within the Cost and Usage Dataset for improved AI reporting. To configure this within CloudHealth, either create and update an existing Data Connect with the location details of a Standard 2.0 export

Dataset Permissions Update

The following permissions for AI datasets have been added to the platform but are not yet enforced. Each read permission corresponds to the standard dataset in the name. These permissions have been added to the Admin and Power user roles by default. Any customers using custom roles will need to add these permissions.

Read AI Cost and Usage

Read AWS AI Cost and Usage

Read Azure AI Cost and Usage

Read GCP AI Cost and Usage

Read OpenAI Cost and Usage

Read Anthropic Cost and Usage

Read MultiProvider AI Cost and Usage

Read GitHub CoPilot Cost and Usage

Read Multicloud Rightsizing Recommendations

As we continue to add more permissions to new datasets and features, customers using custom roles should consider using Restrictive Role Documents. Standard role documents for custom roles operate on an opt-in basis, meaning as new permissions are added to the platform, they must also be added to the custom role document. Restrictive role documents operate on an opt-out basis, customers select which permissions to remove from the role, and all new permissions are added automatically. For customers who have their own versions of Admin, Power User or Standard User, restrictive role documents will allow you to always stay up to date with the latest permissions.

AWS Data Connect in New Experience

Previously only supported by API, it is now possible to manage your AWS Data Connects within the new experience. When using the UI moving forward, all new AWS spend should be configured using an AWS Data Connect. You can find more in our documentation here

As part of this release, CloudHealth automatically migrated all current classic Account configurations for Consolidated and Standalone accounts to Data Connect. Go here for more details on how to migrate to 2.0 as an existing customer.

UI Changes, Updates, and Fixes

• The filter panel now has the top Search section frozen, allowing users to search and click apply while scrolling

• Fix NX Pie chart labels to match dark theme

• NX Assets page: update tab order. Landing tab is first (Inventory) followed by a more logical order of tabs

• When filtering on reports, a loading message now appears when searching large sets of values

• Fixed a scroll issues when cloning widgets inside of dashboard sections

• Improve format of Y axis labels on Cost History

• Users may enable MCP for their tenant in the Settings Page. See the help documentation here.

•  Partners can enable MCP server for  their Channel Customers using GraphQL. 

Deprecation Announcement

1. Deprecation of hourly granularity AWS performance reports.

2. Deprecation of all hourly Archived reports, with the exception of Cost History. 

As of December 1st, 2026, Archived reports with the exception of Cost History across all clouds, and AWS Performance reports will no longer be available. If you have saved reports that meet the above mentioned criteria with hourly granularity, you will need to edit those reports to change the time granularity to daily, weekly or monthly.