This is a control tower for an enterprise agent estate. It governs what agents can do, observes what they are doing, and ensures resilience, compliance, and cost control across the full lifecycle. 

The Agent Ops capabilities are: 

The dashboard provides key monitoring insights to enhance operational efficiency. You can access agent-specific Key Performance Indicators (KPIs), such as the total number of executing agents, metrics on successful and failed executions, and those active for long durations. 

This agent library contains your active agents, making effective management essential. To support this, an observability capability is needed to monitor and manage agents while addressing security and compliance. 

From the Kyndryl Bridge main menu , go to Services → Platform Services → Knowledge & AI → Agent Ops

You can assess tenant health, deployment status, total execution counts, and trends in token usage. 

The dashboard includes an agent topology section displaying model usage, agent status, and a complete agent list. 

The dashboard contains several sections. See the sections detailed below. 

Shows all the components functioning properly. This overview covers a specific time frame, and users can refresh the information to see the latest updates.  

Shows the total number of agents registered during the specified period. Alongside the count, a percentage change indicates whether registrations increased or decreased compared to the previous period. 

The total number of unique deployments that have taken place. Additionally, a percentage change shows any increase or decrease in deployments relative to the last reporting period. 

Shows the cumulative number of executions during the defined timeframe, along with a percentage change indicating whether this figure increased or decreased compared to the previous period. 

Lists the total number of successful executions. It also provides a percentage change to show whether successful outcomes increased or decreased compared to earlier results.  

Reports the total number of executions that failed. For context, a percentage change shows how this figure compares, whether it has increased or decreased, relative to the last reporting period. 

Captures total token usage, including both input and output tokens. This figure is accompanied by a percentage change comparing current usage to the previous period, highlighting any upward or downward trends. 

These are metrics for each agent that encompass execution counts, success rates, and recent statistics, along with their status (active, inactive, or in draft mode) and overall capacity insights. These metrics are presented in a dropdown view, to select the dimensions required. 

You can click the Add Chart button to add more dimensions.  

If you click on any of the dimension’s information, the table automatically filters and displays an Active filter bar to confirm the filter has been applied.

Shows the distribution of AI agents based on their age. 

Offers an in-depth examination of successful and failed executions over the past week. It aims to identify trends and discern patterns within this specific timeframe, providing valuable insights into performance and operational efficiency. 

This table lists active agents and details of essential metrics such as agent name, total executions, success rate, and the date of the last execution. The Agent Ageing column has been implemented to add more context to this table.  

 This section allows you to search, filter and sort any agent deployment activity. It contains the following columns: 

This section provides an overview of the MCP server and the total count of available tools, helping track resources and tool utilization efficiently.