Core Concepts
Celeredge has a small, consistent vocabulary. Learn these eight ideas and the whole product makes sense. (For an A–Z list, see the Glossary.)
The client hierarchy: Prospect → Account → Engagement
- A Prospect is a potential client you're pursuing.
- When they become a client, the prospect becomes an Account — the client company.
- Each account has one or more Engagements — a specific, scoped piece of work (e.g. "Q3 Operating Model Review"). The engagement is the unit almost everything else attaches to: evidence, assessments, gaps, workstreams, and deliverables all belong to an engagement.
The lifecycle: Diagnose → Plan → Execute → Deliver
Every engagement moves through four phases. A phase rail at the top of the engagement keeps the whole team oriented on where the work is.
| Phase | What happens | Read more |
|---|---|---|
| Diagnose | Assess the client against a framework; surface gaps | Diagnose |
| Plan | Convert gaps into committed workstreams and a roadmap | Plan |
| Execute | Track delivery of workstreams; capture insights | Execute |
| Deliver | Produce client artifacts: reports, decks, dashboards | Deliver |
A persistent lineage strip runs through the engagement, linking the active assessment → the gaps it found → the workstreams that address them → the deliverables that ship. This is how a finished slide can be traced all the way back to the evidence that justified it.
Practices and Frameworks
- A Practice is a methodology area your firm offers — for example Operations, Strategy & Transformation, Cybersecurity, Data & AI, Finance & Risk, or General. Practices are how the firm organizes its expertise.
- A Framework is a structured assessment that lives inside a practice — for example an "Operating Model Assessment" inside Operations. A framework is a tree:
Frameworks are authored and published centrally by Celeredge, then made available to your firm. Your firm controls which practices and frameworks are visible. See Frameworks.
Evidence and grounding
The thing that makes Celeredge trustworthy: AI outputs are grounded. Instead of generating prose from thin air, the AI retrieves real source material —
- chunks of the client's uploaded documents,
- results from connected databases,
- signals and insights already captured on the engagement —
and cites each finding back to that source. Accepted findings are saved as Evidence Records into the engagement's Evidence Vault, so the chain of reasoning is auditable. This is covered in depth in Citations & Trust.
Signals
A Signal is an AI-detected observation from the engagement's data that may warrant action — a risk, an opportunity, an anomaly. Each signal comes with a verbatim evidence quote and a link to its source, and can be turned into an assignable action. See Signals.
Workstreams
A Workstream is a committed piece of delivery work created in the Plan phase from an accepted gap. It carries a lineage chip back to the gap, dimension, and practice it came from, so delivery never loses its "why".
Ontology
Each engagement is mapped to an industry ontology — a model of the client's industry as capabilities, processes, and KPIs (e.g. Food Manufacturing → capabilities → processes → KPIs). The ontology seeds the engagement's KPI tree and helps map frameworks to the client's reality. See Ontology & KPI Tree.
How it all fits together
Next: see who can do what in User Roles.