| Primary role | Research, reason, create, and execute work | Ground AI in the firm's operating knowledge and govern its use |
| What it learns | Context within the tool, account, bot, project, or connected workspace | People, projects, decisions, relationships, commitments, lessons, workflows, corrections, and judgment |
| Where knowledge builds | Primarily inside the selected vendor's environment | In a firm-owned Knowledge System that remains consistent as tools and models change |
| How work is governed | The controls, memory, permissions, and integrations available in each tool | Approved sources, identities, permissions, destinations, writeback controls, and human review |
| How adoption happens | The firm configures each tool and builds its own adoption practices | Implementation, training, adoption coaching, workflow discovery, customer success, and ongoing custody of the underlying Knowledge System |