MCP Server

Use natural language to analyze cloud costs from AI assistants

With the DigiUsher MCP Server, you can analyze your cloud costs in plain English. You can find anomalies and savings from Claude and from other AI assistants.

What is the DigiUsher MCP Server?

The DigiUsher MCP Server connects AI assistants such as Claude to the DigiUsher Technology Value Realization platform. It uses the Model Context Protocol (MCP), an open standard that gives an AI assistant access to external data. The server gives your assistant read-only access to your cloud cost data. You can then ask questions in plain English instead of opening dashboards.

Every tool is read-only. The server never changes your cloud infrastructure or your DigiUsher configuration.

What you can do

  • Query and analyze the spend of every data source that your organization connects. These sources include AWS, Azure, GCP, OCI, Alibaba Cloud, Kubernetes, Databricks, Snowflake, Google Workspace, GitHub, and AI platforms.
  • Find cost anomalies and examine them. An anomaly is a spike, a pattern deviation, or an inefficient resource.
  • Find cost optimization recommendations with an estimate of the possible savings.
  • Track FinOps KPIs over time, such as the savings rate, the commitment coverage, and the cost optimization index.
  • Forecast future costs, with projected totals, trend direction, and confidence ranges.
  • Analyze the chargeback and showback allocations of your teams and projects.
  • Calculate unit economics, which is the cost for one unit of business value.
  • Read your tagging governance policies and the tagging violations.
  • Attribute AI spend and coding-agent spend to agent runs and work items.
  • Explore the resource inventory and the connected data sources.

Prerequisites

Support

On this page