Role Assistants turn a Business Role into governed AI-assisted work — Sales Assistant, Project Assistant, Finance Assistant — that helps an identified person complete recurring work using approved enterprise context, governed capabilities, and Evidence.
This is the third pillar of AI Fabrix. Enterprise Knowledge supplies business context; Operational Trust applies boundaries; Role Assistants perform governed work; Evidence Fabrix proves the outcome. A Role Assistant is human-owned and role-scoped: AI can recommend, prepare, and request work, but it cannot grant itself authority. For the full definition — what a Role Assistant is, what it is not, and the candidate → validate → available lifecycle — see What is a Role Assistant?
Find your question
| Question | Page |
|---|---|
| What is a Role Assistant, and what is it not? | What is a Role Assistant? |
| Why business context instead of prompt engineering? | Business context replaces prompts |
| How is this different from an agent assembled from tools? | Governed capabilities, not agents |
| What is a task, and what outcomes can it have? | Tasks and outcomes |
| How does a request actually pass through governance? | Governed work execution |
| How do skill levels and promotion work? | Skill levels and promotion |
| How do people reach an assistant? | Assistant channels |
| How are Role Assistants created? | How Role Assistants are created |
| What assistants exist today? | Role Assistant examples |
| How do Evidence and governed execution fit together? | Role Assistants, Evidence, and governed execution |
| What is the difference between conversation and Work? | Conversation-first work |
| What is the difference between Ask and Approval? | Human tasks, approvals, and Why |
| How does this work with ChatGPT, Copilot, or external AI over MCP? | External AI and MCP |
| What is the Business Transformation Assistant? | Business Transformation Assistant |
| How do assistants react when business reality changes? | When business reality changes |
| What are the lifecycle states (Draft, Active, Suspended)? | Role Assistant status and availability |
| How do assistants get their capabilities? | How Role Assistants get capabilities |
| How do you prove a Role Assistant works before release? | Proving Role Assistants |
Business value
Role Assistants help organisations move from generic AI interaction to repeatable business work: candidates are generated automatically from Business Roles and approved enterprise definitions, the Business Context Map is reused instead of rebuilt per use case, assistants are tested before release, human decision points stay explicit, and Evidence improves behavior without silently changing what was released.
One-line summary
Role Assistants automatically turn a Business Role and approved business foundation into a candidate governed capability that people validate before use.