Machine interface to AIM decisions
AIM’s path for an organization’s own AI to start and advance decision work — without becoming the signer. This is an additional AIM capability. It appears in a workspace when it is included in that organization’s agreement.
Why this exists
More work will begin in an organization’s own AI — an assistant a person already uses, or an internal orchestrator — rather than by clicking through a software screen. If AIM only works as a human workspace, that work goes around AIM. The decision record, the scoring, and the audit trail never form.
Agent Orchestration exists so an approved interface can open an assessment, run analysis, and draft documents inside AIM’s existing workflow. People still accept the stack, send work for review, complete the assessment, and attest the output. AIM remains the system of decision authority.
What it does
An approved interface can start or advance an assessment, run AIM’s analysis engines, draft modernization documents, and propose a next step for a person to review. That work lands in the same assessment workspace people already use.
What stays with people
This is separate from Mission Control. An approved interface drafts and advances decision work inside AIM. Eagle remains the in-product assistant. Completing, awarding, attesting, and voting stay with people. Changing a live environment stays with the organization's own approval path.
How AIM governs it
An approved machine interface is not a person signed into AIM. It cannot borrow a browser session or act as a named reviewer.
Work stays inside the organization that authorized it. Another organization's assessments, documents, and records are not visible.
Completing an assessment, recording an award, attesting a document, and voting on a change remain human actions.
Text submitted through this interface is inspected before it is stored or used in AIM analysis.
When an interface drafts or proposes a step, people review it in AIM. Nothing silent-completes the record.
Assessments, scoring, architecture, and documents remain fully usable with no machine interface configured.
Two different capabilities
An organization’s AI works inside AIM’s decision workflow: assessments, analysis, drafts, and proposed next steps. People still finalize irreversible decisions.
AIM observes the live environment against the documented record and packages reviewable work. Changing a live environment stays with the organization's own approval path.
Both are additional AIM capabilities. Neither is required for an assessment. Including one in an agreement does not include the other.
Related reading
Observation, review packages, and no-change rehearsals in the customer environment.
How AIM keeps an auditable record of AI-assisted decision work.
Where AIM uses deterministic engines first, and how models stay bounded.
How organizational information is collected, isolated, and retained.