How ConvergeSense fits together
A short guide to the pieces between your agent and your infrastructure.
ConvergeSense connects AI agents to reviewed operations across your infrastructure, monitoring, and ticketing systems. The platform coordinates access; an onsite Engine talks to the systems it can reach.
Workspaces
A workspace is the tenant boundary for your team, systems, and policies. Agent OAuth consent selects exactly one workspace. Membership, your role, the integration grant, and tool policy determine the operations available to you.
Engines
An Engine is an onsite runtime. It connects outbound to the control plane, receives approved work, and returns bounded, normalized results. Equipment credentials and raw terminal or API sessions remain onsite.
Start with Engine setup.
Connectors and connections
A connector describes reviewed operations for a supported system. A connection binds a particular system to an Engine and an exact connector release. Testing and activation establish whether that binding is ready.
Connector and contract digests identify the version being used. A changed contract can make an existing tool unready until it is reviewed again. See Connectors.
Tools
A tool exposes a specific approved operation to your agent. Its contract describes the accepted inputs and bounded result. Discovery includes only tools available to your current workspace, role, and client.
An agent searches before selecting a tool, describes its contract before running it, and keeps the same job ID while waiting for results. See Tools and permissions.
Evidence and freshness
Results carry observation times and freshness information. A cached observation can still be useful, but it should not be presented as a new live measurement. Warnings and incomplete coverage are part of the answer.
Workflows and playbooks
A workflow combines reviewed steps into a published sequence. Operational playbooks add investigation guidance and structured authoring when enabled for the workspace. A saved draft does not publish or schedule itself.
See Operational playbooks for the review and execution flow.