A professional services engagement
The question people ask: what did we agree, what did the client decide in which session, what is due when, and what changed since the statement of work?
What feeds the graph: the statement of work, session notes, decision logs, deliverable reviews, steering committee minutes, and what was said in the room.
What the graph holds
| Class | For example |
|---|---|
Engagement, Workstream, Deliverable |
the claims transformation, its operating-model workstream, the target operating model |
Session, Decision, ActionItem, Question |
the workshop of 2026-05-12, the decision to keep the legacy intake, the action to confirm it, the question it answers |
Organization, Person, Role, GovernanceForum |
the client and the firm, the sponsor, the role they fill, the steering committee |
Concept, Framework, Document |
a term the engagement defined, the method it follows, the minutes that record all of it |
A session raised a decision and is recorded_in the minutes; a decision decides a deliverable or a concept, an action item implements it and answers a question; a person fills a role and owns a workstream; a later decision supersedes an earlier one. The graph is the engagement's memory: what the client decided, when, and where it is written.
A cited answer
oto query --project engagement entity "Keep the legacy intake"
oto query --project engagement by-type Decision
oto query --project engagement stale
A decision's card names the session that raised it, the date, and the minutes it is recorded in. stale lists the decisions that were later reversed, and by what.
How it is used
- The engagement lead captures decisions at the end of each session with the capture skill; the graph supersedes what changed.
- A consultant joining in month three asks the agent what was decided about intake and gets the decision, the session and the minutes.
- The client sees the same graph through a view, without the working documents behind it.
- The next engagement starts from an ontology exported from this one, with its reasoning and without its data.
The commands and the skills
oto init --name "Meridian Transformation" --ontology professional-services --project engagement
oto capture --project engagement --title "Steering committee" --by "<sponsor, role>" --at 2026-05-12 --statement "..."
oto ontology export --project engagement --name firm-engagement --summary "Engagements, as the firm runs them"
Skills: capture, curate, spec for the deliverables written from the graph. Tutorial: your vocabulary becomes a pack.