> For the complete documentation index, see [llms.txt](https://developer.harness.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://developer.harness.io/roadmap/cloud-and-ai-cost-management.md).

# Cloud & AI Cost Management

Explore planned Cloud and AI Cost Management capabilities and enhancements.

This page tracks planned capabilities and enhancements for Cloud and AI Cost Management.

### Released

What has been released

<table data-column-title-hidden data-view="cards" data-search="true"><thead><tr><th>Tag</th><th>Title</th><th>Description</th></tr></thead><tbody><tr><td>Cost Visibility</td><td><strong>AI-assisted Report creation</strong></td><td>AI-powered creation of Perspectives/Views including discovery of relevant cost categories, labels, tags, and rule creation.</td></tr><tr><td>Cost Visibility</td><td><strong>AI-assisted Unit Costs creation</strong></td><td>AI-powered creation of unit cost definitions including discovery of relevant unit metrics.</td></tr><tr><td>Cost Visibility</td><td><strong>AI-assisted Cost Insights</strong></td><td>AI-powered cost insights and analysis on cloud spend, top spenders, trends, etc across cloud and AI spend.</td></tr><tr><td>Cost Visibility</td><td><strong>AI-assisted Commitment analysis</strong></td><td>AI-powered analysis on commitment portfolio, savings, coverage, utilization and optimization opportunities.</td></tr><tr><td>Cost Visibility</td><td><strong>AI-assisted Cost Categories creation</strong></td><td>AI-powered creation of Cost Categories including discovery of relevant labels and tags, and rule creation.</td></tr><tr><td>AI Cost Management</td><td><strong>AI Traces for Cost Management</strong></td><td>Granular AI Traces for cost management at the agent, service, session, model, run and request level.</td></tr><tr><td>AI Cost Management</td><td><strong>Additional AI Cost Connectors for Cursor, GitHub Co-pilot, Claude Desktop</strong></td><td>Additional AI Cost connectors for cost management of Cursor, GitHub Co-pilot, Claude Desktop.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Anomaly FinOps AI Assistant</strong></td><td>AI-powered insights to help FinOps teams analyze anomalies and notify the right stakeholders.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Whitelisting of RI, SP, Services for anomalies</strong></td><td>Whitelisting of Reserved Instances, Savings Plans, and Services to exclude them while generating anomalies.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Capture Recommendations Savings in Jira and Snow</strong></td><td>Capture recommendations savings in Jira and ServiceNow for tracking and reporting.</td></tr><tr><td>Commitment Orchestrator</td><td><strong>Commitment Orchestrator for RDS GA</strong></td><td>General Availability of Commitment Orchestrator for Amazon RDS, enabling better cost-efficiency through commitment management for database services.</td></tr><tr><td>Commitment Orchestrator</td><td><strong>AWS Savings Plan Database Discovery/Support</strong></td><td>Discovery and support for AWS Savings Plans for RDS and ElastiCache databases.</td></tr><tr><td>Cost Visibility</td><td><strong>CACM Overview Page Revamp</strong></td><td>Revamp of the CACM Overview page.</td></tr><tr><td>Cost Visibility</td><td><strong>Unit Cost Economics</strong></td><td>Unit Cost Economics to track cost per outcome and other complex unit cost definitions, including ingestion of custom unit metrics.</td></tr><tr><td>Cost Visibility</td><td><strong>Data Ingestion Status Update</strong></td><td>Introduces a centralized Data Jobs Status section and real-time notifications for active data jobs, third-party outages, and widget-specific alerts across ingestion, Cost Category backfills, PDT updates, and external issues impacting Harness CACM.</td></tr><tr><td>AI Cost Management</td><td><strong>AI Cost Management</strong></td><td>One attribution model for AI costs. Production AI maps to session and interaction, developer AI maps to PR, commit, and ticket, and both roll up into unit costs.</td></tr><tr><td>Cost Visibility</td><td><strong>Unit Economics</strong></td><td>Track cost per customer, transaction, or feature so finance, product, and engineering can tie spend to business outcomes.</td></tr><tr><td>Commitment Orchestrator</td><td><strong>Commitment Orchestrator Expanded Support</strong></td><td>Automatically purchase, renew, exchange, and rebalance RIs and SPs as usage shifts, now with extended support for RDS, ElastiCache, and OpenSearch.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Auto-Inference of Savings</strong></td><td>Auto-inference of savings (partially realized) in Recommendations.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Budget perspective decouple</strong></td><td>Decoupling budget perspectives with cost category support.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>AWS Passthrough recommendations support with Cost Optimization Hub (EC2, EBS, RDS)</strong></td><td>Increases the scope of Recommendations across all major AWS resource types, giving you a complete view of savings opportunities in your cloud environment.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Anomaly Lookback Support</strong></td><td>[Released: February 2026] Anomaly lookback support to detect anomalies based on historical data.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Anomaly Detection v2 (GA)</strong></td><td>[Released: January 2026] Enhanced anomaly detection capabilities with drill-down analysis. Includes support for cost category anomalies, customizable workflows, and preferences for setting percentage deviations and minimum cost thresholds.</td></tr><tr><td>AutoStopping</td><td><strong>AutoStopping Alerts</strong></td><td>[Released: January 2026] Configure notifications for critical events related to your AutoStopping rules.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Auto-Inference of Savings</strong></td><td>[Released: December 2025] Auto-inference of savings (fully realized) in Recommendations.</td></tr><tr><td>Cluster Orchestrator</td><td><strong>Nodepools and Nodeclasses Configuration via Harness UI</strong></td><td>[Released: December 2026] UI-based configuration management for Nodepools and Nodeclasses.</td></tr><tr><td>Cost Visibility</td><td><strong>Cost Category Enhancements</strong></td><td>[Released: December 2025] Performance improvements - 2x Query speeds, Support to leverage cost categories across all CACM features: Recommendations, Asset Governance, Budgets</td></tr><tr><td>Cluster Orchestrator</td><td><strong>Cluster Orchestrator for EKS</strong></td><td>[Released: December 2025] Workload-driven intelligent node autoscaling with distributed spot orchestration</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Tags support for Recommendations</strong></td><td>[Released: October 2025] Tags are supported for recommendations and filtering is allowed.</td></tr><tr><td>Asset Governance</td><td><strong>FInOps AI assistant (Beta)</strong></td><td>[Released: December 2025] The FinOps AI Assistant leverages generative AI to automate the creation and enforcement of cloud governance policies, enabling faster cloud optimization.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Bulk Management and Filtering in Budgets</strong></td><td>[Released: October 2025] Select and modify multiple budgets simultaneously, adjust budget amounts by percentage or fixed value, manage alerts across multiple budgets (add/remove recipients, delete alerts), clone budgets with customizable options for thresholds and recipients</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Ticketing Tool Mapping</strong></td><td>[Released: October 2025] Ticketing Tool Mapping for Recommendations</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Budget Folders</strong></td><td>[Released: October 2025] Budget Folders to improve usability and streamline budget management. Budgets are now automatically associated with the Perspective folder under which they were created. This provides a consistent folder structure across both Perspectives and Budgets.</td></tr><tr><td>Cost Visibility</td><td><strong>Dynamic Cost Categories Toggle in Perspectives</strong></td><td>[Released: September 2025] Dynamic Cost Categories Toggle in Perspectives</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Jira Status Mapping in Recommendations</strong></td><td>[Released: August 2025] Jira Status Mapping in Recommendations</td></tr><tr><td>Cluster Orchestrator</td><td><strong>Cluster Capacity Limits</strong></td><td>[Released: August 2025]For Karpenter Nodepools, users can now set maximum CPU (cores) and memory (bytes) limits as guardrails to prevent uncontrolled cluster scaling.</td></tr><tr><td>Asset Governance</td><td><strong>Rules Generating Recommendations</strong></td><td>[Released: August 2025] Rule and Target Account Exclusions in Governance Recommendations, enabling you to define custom default rules that apply globally or to specific account subsets.</td></tr><tr><td>Cost Visibility</td><td><strong>Dynamic Perspective Reports</strong></td><td>[Released: August 2025]Create reports from your perspectives to bookmark specific filter and grouping configurations. No need to rebuild the same view repeatedly — just save it once and access it anytime.</td></tr><tr><td>Commitment Orchestrator</td><td><strong>Commitment Orchestrator Inventory</strong></td><td>[Released: September 2025] Unified inventory eliminates the need for separate AWS views or manual exports, giving you complete visibility directly inside Harness.</td></tr><tr><td>Cluster Orchestrator</td><td><strong>Cluster Orchestrator Schedules</strong></td><td>[Released: December 2025] Cluster Orchestrator schedules allowing users to define usage windows for shared resources.</td></tr><tr><td>Cluster Orchestrator</td><td><strong>VPA Support for Cluster Orchestrator</strong></td><td>[Released: November 2025] Automatically adjusts the CPU and memory resource requests and limits of containers in pods based on their historical usage</td></tr><tr><td>Commitment Orchestrator</td><td><strong>RDS (Beta)</strong></td><td>Beta release of Commitment Orchestrator for Amazon RDS, enabling better cost-efficiency through commitment management for database services.</td></tr></tbody></table>

### Now

Q3 CY'26, Aug 2026 - Oct 2026

<table data-column-title-hidden data-view="cards" data-search="true"><thead><tr><th>Tag</th><th>Title</th><th>Description</th></tr></thead><tbody><tr><td>Commitment Orchestrator</td><td><strong>Set Advance Email Alerts for Expiring RIs and SPs</strong></td><td>Sends proactive email alerts when Reserved Instances (RIs) and Savings Plans (SPs) are nearing expiration, so customers can renew or adjust before losing coverage and incurring unexpected on-demand charges.</td></tr><tr><td>AutoStopping</td><td><strong>Azure &#x26; AWS Spot Support in AutoStopping</strong></td><td>Adds support for Azure and AWS Spot instances in AutoStopping to optimize cost savings.</td></tr><tr><td>Cost Visibility</td><td><strong>K8S Data Ingestion (SCAD)</strong></td><td>Improves Kubernetes cost data ingestion accuracy and granularity, addressing lost events, node-status discrepancies, and cost visibility for delegate-less K8s clusters.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Data Scope Support Extended to Recommendations, Anomalies &#x26; Governance</strong></td><td>Extends enterprise Data Scope access controls to Recommendations, Anomalies, and Governance, so restricted users only see recommendations/anomalies within their permitted scope.</td></tr><tr><td>Commitment Orchestrator</td><td><strong>Cluster &#x26; Commitment Integration Enhancements</strong></td><td>Improves how Cluster Orchestrator savings and Commitment Orchestrator coverage work together, giving a more complete view of cluster cost savings.</td></tr><tr><td>Cost Visibility</td><td><strong>Unit Cost Economics Support in Custom Dashboards</strong></td><td>Unit cost economics and ingested unit metrics for cost per outcome within Custom Dashboards.</td></tr><tr><td>AutoStopping</td><td><strong>Enhanced AutoStopping Terraform Support</strong></td><td>Enhances Terraform support for configuring AutoStopping via Infrastructure as Code with drift detection.</td></tr><tr><td>Commitment Orchestrator</td><td><strong>Azure VM Support for Commitment Orchestration</strong></td><td>Granular visibility and orchestration of Azure VM committments.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Enhanced Recommendations Experience</strong></td><td>Enhanced experience reporting and usability for optimization Recommendations.</td></tr><tr><td>Cost Visibility</td><td><strong>Shift-left actioning of Recommendations with IaCM Integration</strong></td><td>Integration of Infrastructure-as-Code Management (IaCM) to surface resource level recommendations in workspaces and auto apply recommendations at source without drift.</td></tr><tr><td>AI Cost Management</td><td><strong>User AI Cost Governance</strong></td><td>Active governance and enforcement of user-level AI spend with enforcement at the user, provider and model level.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Azure Passthrough recommendations support with Azure Advisor</strong></td><td>Support for Azure Passthrough recommendations for: Snapshot to Standard, VM Reserved Instances, Blob Reserved Capacity, Delete Unattached Disks, Files Reserved Capacity.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Anomaly detection integration with JIRA and ServiceNow</strong></td><td>First class integration with JIRA and ServiceNow to raise tickets and delegate actioning of detected anomalies.</td></tr><tr><td>Commitment Orchestrator</td><td><strong>Support RIs for OpenSearch with Commitment Orchestrator</strong></td><td>Support RIs for OpenSearch in Commitment Orchestrator.</td></tr><tr><td>AI Cost Management</td><td><strong>AI Spend Explorer by Users</strong></td><td>Breaks down AI spend by individual user, so teams can see who is driving AI cost.</td></tr><tr><td>AI Cost Management</td><td><strong>AI Spend Explorer by Agents</strong></td><td>Breaks down AI spend by agent/model, so teams can see which AI agents are driving cost.</td></tr><tr><td>AI Cost Management</td><td><strong>AI Spend Capitalization Insights</strong></td><td>Surfaces which AI spend can be capitalized vs expensed, helping finance teams with accounting treatment of AI costs.</td></tr><tr><td>AI Cost Management</td><td><strong>Engineering ROI Intelligence</strong></td><td>Connects engineering spend (including AI) to output and business impact, giving a view into engineering return on investment.</td></tr></tbody></table>

### Later

Q4 CY'26 and beyond

<table data-column-title-hidden data-view="cards" data-search="true"><thead><tr><th>Tag</th><th>Title</th><th>Description</th></tr></thead><tbody><tr><td>Commitment Orchestrator</td><td><strong>ElastiCache Commitment Support</strong></td><td>Commitment Orchestrator can discover and purchase Reserved Instances for ElastiCache usage.</td></tr><tr><td>Cluster Orchestrator</td><td><strong>Cluster Orchestrator for AKS</strong></td><td>Cluster Orchestrator support for Azure Kubernetes Service (AKS), including discovery, scale up/down, Spot handling (interruption, fallback, reverse fallback, schedule windows), Karpenter (latest Azure version), Nodepool/Nodeclass setup, and policy-driven automated provisioning and node pool management.</td></tr><tr><td>Commitment Orchestrator</td><td><strong>Reject Lifecycle for Commitment Orchestrator Recommendation</strong></td><td>Lets customers reject/dismiss Commitment Orchestrator recommendations they do not want to act on.</td></tr><tr><td>AI Cost Management</td><td><strong>AI Cost Optimization Recommendations</strong></td><td>Optimization recommendations for AI spend with model routing, context window optimization, prompt caching, etc.</td></tr><tr><td>AI Cost Management</td><td><strong>Production Agent AI Cost Governance</strong></td><td>Active governance and enforcement of production agent AI spend with enforcement at the agent, provider and model level.</td></tr><tr><td>AutoStopping</td><td><strong>Smart Advisor for Kubernetes Cluster</strong></td><td>Smart Advisor helps you uncover cost-saving opportunities by identifying underused cloud resources and recommending the ideal Autostopping configurations.</td></tr><tr><td>AutoStopping</td><td><strong>Autostopping Global Schedules</strong></td><td>Global schedules for Autostopping allowing users to define usage windows for shared resources.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>GCP Passthrough recommendations Support with Google Cloud Recommender</strong></td><td>Support for GCP Passthrough recommendations with Google Cloud Recommender to increase the scope of recommendations across all major GCP resource types.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Dynamic Anomaly Detection thresholds</strong></td><td>Intelligent threshold setting for anomalies at account and cost category level.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Recommendations OPA Policy - CD integration</strong></td><td>Recommendations OPA Policy with CD integration.</td></tr><tr><td>BAR (Budgets, Anomalies, Recommendations)</td><td><strong>Show impact of applied recommendations using perspectives or reports</strong></td><td>Show impact of applied recommendations using perspectives and reports.</td></tr><tr><td>AI Cost Management</td><td><strong>AI Cost &#x26; Performance Benchmarking</strong></td><td>Benchmarks AI cost and performance across models/providers to help teams choose the most cost-effective option for their workload.</td></tr><tr><td>AI Cost Management</td><td><strong>Context-Aware Engineering Governance</strong></td><td>Applies governance policies to engineering/AI usage based on context (e.g. project, team, environment) rather than static rules.</td></tr></tbody></table>

{% @harness-feedback/feedback module="roadmap" pagePath="roadmap/cloud-and-ai-cost-management" %}