Ask, Analyse, Identify, Classify and Compare. Every result shows its source, its confidence and the next human action, and every result can start a workflow. Intelligence is the understanding layer of the DocxIntel workspace, and it runs inside your infrastructure.
Accuracy is measured field-by-field against human-adjudicated ground truth; the document mix and methodology are published on the accuracy page.
Ask across the documents you are permitted to see. Every substantive answer exposes its source documents and evidence; if the answer cannot be supported, the result says so.
Structured analysis jobs return findings grouped with evidence, severity, confidence, affected section or page, recommended action and workflow action.
Extracted fields and entities in a structured table with source location, confidence and editable values. Corrections are stored as feedback and audit events.
Classification returns the document type and category with confidence, evidence and recommended routing. Bulk classification assigns work to department queues.
Additions, removals, changed values, numerical changes, clause deviations and policy impact — side by side with evidence and a one-click review workflow.
Renewal, expiry, probation and obligation dates found in documents become monitored work items, with owners alerted before anything is missed.
A finding, a field, a classification or a difference — each with source, confidence and status.
Confirm a field, review a finding, accept a route — the item lands in the right person’s My Work.
One click turns a deviation or a risk into a review, an approval or an exception case.
The result, the decision and who made it are recorded — for people and for the Agent.
What you get
Evidence
When unsure
What happens next
Audit
Where it runs
Straight answers, no sales deck. If the answer you need is missing, ask us directly.
Each Intelligence capability returns a structured result tied to a document-work object — a finding, a field, a classification, a difference — with its evidence, its confidence and the human action that comes next. Results feed workflows, approvals and the audit log; they are not answers that stop at the chat window.
