Understanding Standpoint's evidence & review workflow

Who this page is for

This page is for anyone reviewing a Carbonbit Standpoint client's evidence as part of an audit or assurance engagement — Carbonbit's own internal assurance staff, or an external, independent assessor. It explains how the system itself works: where evidence lives, what the platform's own grading tags mean, and the shape of the review workflow.

It does not, and cannot, tell you anything about a specific client, engagement, or piece of evidence — this page has no way to know who is reading it or which engagement you are reviewing. For a question about a real client's actual position, evidence, or record, that always goes to the named consultant or the Carbonbit team responsible for that engagement, never to this page or to the platform's own AI assistants.

How evidence gets into the system

A client's evidence — invoices, certificates, policy documents, energy audit reports and similar — enters Standpoint through a small number of defined channels: a document upload, an automated data sync (for example, an accounting system connection), or a structured intake link sent to a named contributor for one specific request.

Nothing entered this way becomes part of a client's approved record automatically. Every item passes through an extraction and review step first — the system reads what it can from the document, and a human reviewer confirms, corrects, or rejects that reading before it is approved. There is no path that lets a document skip this review, whatever channel it arrived through.

The grading ladder — what a provenance tag actually means

Every approved figure on the platform carries a grade describing how it was obtained, shown as a small tag wherever the figure appears. From weakest to strongest: self-reported (typed in directly, no supporting document), document-evidenced (backed by an uploaded document a reviewer has checked), system-extracted (read automatically from a connected system, such as an accounting integration), and third-party-verified (confirmed by an independent verifier acting in that specific role, segregated from whoever prepared the record).

A grade can only move up this ladder, and only through a real event — a document being reviewed, a system connection being confirmed, an independent verifier signing off. It is never silently upgraded, and a figure is never presented as more strongly evidenced than the grade on file actually shows.

Independent verification specifically requires a person acting in a segregated verifier role on that engagement — someone who also prepared the client's own submissions cannot verify them. That segregation is enforced by the platform itself, not left to a reviewer's own discipline.

What travels with a number

Alongside its grade, a material figure carries where it came from (the source document or record), how it was produced (the extraction or calculation method, and which version of any factor set was applied), when it entered the system and when it was approved, and who approved it — a named reviewer, or the specific automated rule that admitted it. A figure missing any of this is not treated as approved data.

Records are never edited in place once approved. A correction or an update creates a new version; the record it replaces stays on file, so the history of what was believed and when remains reconstructable, not overwritten.

The shape of the review workflow

Most evidence — a policy document, a certificate, an extracted figure — moves through the same basic shape: submitted, held in a pending-review state, and then approved or rejected by a reviewer, with any rejection reason recorded alongside it. A policy document specifically also carries its stated document version, a named internal owner, and its next review date.

A rejected or corrected item is not deleted — the record of what was submitted and why it did not clear review stays on file, the same append-only discipline as everything else here.

What a consultant or client would typically produce for a review

A consultant preparing for a review typically works from the client's own approved record — the same evidence, grades and provenance described above, filtered to what a given review actually needs. There is no separate, review-specific copy of the data that could drift from what the platform itself shows day to day; the evidence a reviewer sees is the same evidence the client's own outputs are built from.

If you need something this page cannot give you

This page is deliberately generic, static, and unable to answer anything about a real engagement. If you need to review actual evidence for a specific client, or have a question about a specific record, contact the named consultant or the Carbonbit team responsible for that engagement.