Docs/Building with Celiq/Business Knowledge Graph

Business Knowledge Graph

Connect the entities that matter for business decisions — People, Teams, Goals, KPIs, Actions, Experiments, Results. Orion answers like an employee who knows the whole context.

WeaverLuminaryUpdated Jun 2026 · 5 min read

✦

In this section

Part of Celiq's semantic layer platform. Connect your warehouse, model your data once, query it everywhere.

💡

Tip

What it is: A graph of your business context — people, experiments, OKRs, and relationships — that makes Orion answers dramatically more specific.
Why use it: Without it, Orion answers "CAC increased" — with it, Orion answers "CAC increased after the email campaign paused, owned by Ravi, which aligns with your Q3 CAC target."
Where to find it: Suite → AI Analysts → Knowledge Graph (/knowledge)
Who can use it: Weaver and Luminary roles can add and edit. All roles benefit from the improved answers.

Business Knowledge Graph

Connect the entities that matter for business decisions: People → Teams → Goals → KPIs → Actions → Experiments → Results. When this graph exists, Orion answers like an employee who knows the whole context — not just a query engine.

Accessing the Knowledge Graph

Suite → AI Analysts → Knowledge Graph (/knowledge)

What to add

People

Add team members and which metrics they own. This lets Orion say "CAC is owned by Ravi (Marketing)" in its answers.

Experiments

Log A/B tests, campaigns, and initiatives with:

Name and description
Start date and end date
Owner
Outcome (once available)

This lets Orion connect metric movements to the experiments that may have caused them.

OKRs

Add objectives and key results by quarter. Orion uses these to put individual metric movements in strategic context — "revenue is 12% below the Q3 OKR" rather than just "revenue is down."

How it improves Orion answers

Before knowledge graph:

"CAC increased by 18% last week."

After knowledge graph:

"CAC dropped after the email campaign launched (Jun 3, owner: Ravi). This aligns with your Q3 CAC target of <$40. The campaign experiment is currently active and showing positive early results."

The graph gives Orion the "who", "why", and "so what" for every metric movement.

Testing the context

Use the Context Preview tab to see exactly what context Orion would receive for any question. This is useful for verifying that your knowledge graph entries are being incorporated correctly.

What happens if the graph is empty

Orion works without the knowledge graph — it will still answer data questions accurately from your semantic layer. The knowledge graph is additive: everything in the graph improves answers, nothing breaks without it.