A structured record of every individual operating within or on behalf of the organization: who they are, what they are authorized to do, what they are capable of, who they work with, and every system where their data lives.
Every organization has people data. The problem is that it lives everywhere at once. A person's record is split across an HRMS, a project management tool, an access provisioning system, an org chart, a communication platform, and in many cases a spreadsheet someone maintains because nothing else has the full picture. Each system knows something different. None of them know everything. And none of them talk to each other. The Actor Index creates the authoritative layer that knows what each system holds and where to find it. When someone needs information about a person, the index tells them exactly where to look. For AI, this is a prerequisite. An agent that needs to route work, assign tasks, assess capacity, or respect authority boundaries is reading from this index. Without it, the agent is guessing or interrupting a human every time it hits a gap.
The Actor Index is the first dimension of the Awareness phase in the Ragsdale Framework for Autonomization, developed by Marc Ragsdale. Prospus implements it as part of structured AI transformation engagements. Kaamfu operationalizes it as a native layer inside the Autonomous Operating Environment, making actor data queryable by AI agents across the platform.
Actor data is fragmented across systems that do not agree. Nobody has a complete picture of who owns what. Role changes update in one system and nowhere else. When someone leaves, their files, access, and context go with them. Leadership cannot answer who is responsible without asking around. AI cannot act on people it cannot find.
Every actor has a single authoritative record that is current and queryable. Roles, authority, capabilities, and assignments are defined and visible. When someone joins, the index defines what needs to be created. When someone leaves, nothing walks out the door. AI agents can route, assign, and govern around people with confidence.
| Metric | Before | After |
|---|---|---|
| Visibility | Leadership assembles a picture by asking people, waiting, and reconciling conflicting versions from different systems. | Every actor has a single authoritative record. Leadership can answer who owns what without initiating a search. |
| Update Mechanism | Role changes update in one system and nowhere else. A promoted employee retains prior-role access permissions for months. The conflict surfaces only when an audit runs. |
Role changes trigger updates across the index. One record reflects current state. |
| Completeness | Each system holds a fragment. HR knows the contract. The project tool knows the assignments. Provisioning knows the access. Nobody knows all three. A new manager asks what their direct report is working on. Three systems, two manual exports, one spreadsheet later, still without a clear answer. |
The index holds the complete actor record and routes to source systems for full detail. |
| Queryability | You navigate systems to find people and hope the right system has what you need. Who has authority to approve expenses over $10k? Ask around. Expect two days. |
Natural language query returns the actor record and routes to source systems. The answer arrives in seconds. |
| Ownership Model | Actor data lives in individual accounts and personal folders. A departing employee takes context with them. A senior account manager leaves. Their client notes are in a personal Google Drive. The folder disappears inside 48 hours. |
All actor data is organizationally owned. Access is governed. Offboarding does not create loss. |
| Offboarding | Knowledge walks out. Files, relationships, and institutional memory disappear with the person. A developer leaves mid-project. Nobody knows what systems they had access to or what they were working on. |
Structured handoff. Everything the person owned is documented and transferred before departure. |
| Authority Clarity | Authority is assumed, informal, or buried in a job description nobody has read. A vendor needs contract approval. Three people think they have authority. One does. The project waits four days to find out. |
Every actor's decision authority is explicit, current, and queryable. AI governance decisions reference it directly. |
A score of 10 on the Actor Index means this dimension is fully resolved and no longer a constraint on the phases that follow. Here is what that requires in practice.
Inventory every system that holds actor data: HRMS, project tools, provisioning, directories, shared drives, contracts, and any spreadsheet someone maintains. The goal is a complete picture of what exists and where before any consolidation decisions are made.
Assess each system against operational requirements. Does it reflect current state in real time? Is it updated manually or automatically? Is the data organizationally owned or held in individual accounts? This assessment produces a clear picture of where the stack falls short.
Define the schema and build the authoritative layer. Agree on the fields the index must hold for every actor. Build the index as a structured knowledge layer that references source systems rather than replicating them.
Assign ownership and establish the update cadence. Define the trigger: new hire, role change, departure, project transition. Without an explicit cadence and owner, the index decays within weeks.
Make the index queryable. Inside Kaamfu for organizations on the platform, or as a standalone natural language interface for those that are not.
They build a directory and call it an Actor Index. A directory lists people. An index defines what each person means to the organization, and that distinction requires conversations most teams would rather avoid. The authority and accountability fields are the ones that create friction. They are also the ones that make the index worth building.
The Autonomy Diagnostic scores the Actor Index on a 0 to 10 scale. Each point reflects a specific observable state in your organization.
No structured record of organizational actors exists. Who owns what is determined by asking around.
An employee directory exists with names and titles. No authority, capability, or assignment data is captured.
Actor data exists across multiple systems but conflicts between them are common and no single source is authoritative.
A primary system of record exists for employment data but role authority, capabilities, and assignments are not consolidated.
Actor records cover identity and reporting structure. Authority and capabilities are partially documented but not consistently maintained.
A structured actor record exists covering most fields. Update cadence is defined but not consistently followed after role changes.
The index covers identity, role, authority, and assignments for most actors. Conflicts between source systems are visible and being resolved.
The index is complete for all employees and key contractors. Update cadence is functioning. Offboarding procedures reference the index.
The index covers all actors including vendors and embedded contractors. It is queryable and referenced by AI agents for routing and assignment decisions.
Every actor has a complete, current, organizationally-owned record. Authority, capabilities, and assignments are queryable in natural language. Onboarding and offboarding are governed by the index. AI agents route and govern from it without human intervention.
No. An HRMS holds employment records. The Actor Index holds the operational picture: what each person is authorized to do, what they are capable of, what they have agreed to, and where their data lives across every system. The HRMS is one of several source systems the index references, not the index itself.
No. The Actor Index is a reference layer. The HRMS still holds employment records. The project tool still holds assignments. The index knows what each holds, routes to it, and resolves conflicts between them. Migration decisions come later, grounded in a clear picture of what the organization actually runs on.
Accuracy is a governance problem, not a technical one. The index needs an owner and a defined trigger for updates: new hire, role change, project transition, departure. Without an explicit cadence, any index decays within weeks.
AI agents that assign work, route requests, assess capacity, or enforce authority boundaries are reading from this index, or asking a human to fill the gap every time. The index is what makes autonomous operation around people possible.
Access is governed at the index level. Compensation is available to those with explicit access: finance, HR leadership, and executive roles. It is not surfaced in general queries.
The actor record is not deleted. It is archived. The System Footprint tells the organization exactly what accounts and access need to be revoked. The record remains queryable for historical and audit purposes.
Yes. Anyone operating within or on behalf of the organization belongs in the Actor Index: employees, contractors, embedded vendors, and any external party with ongoing access or accountability.
The index surfaces conflicts rather than silently resolving them. When the HRMS lists one title and the provisioning system lists another, the index flags that conflict and routes it to the record owner for resolution.
An org chart shows reporting lines. The Actor Index holds everything the org chart cannot: authority levels, capabilities, compensation, agreements, system access, and the full operational picture of what each person does and is accountable for.
A score of 7 or above means the Actor Index is functioning well enough that it is not a meaningful constraint on the phases that follow. The Autonomy Diagnostic will tell you which dimensions need the most attention first.
The Autonomy Diagnostic scores every dimension of the Ragsdale Framework and tells you exactly where to focus first.
Take the Diagnostic