Skip to main content
After the AI qualification stage scores an inbound lead, no action happens automatically. Every lead waits for a human decision. Operator Review is that checkpoint — a protected dashboard where authenticated team members review the AI’s assessment alongside the original lead data and record a decision that moves the lead forward or removes it from the pipeline. This page explains what operators see, the decisions available to them, and how every action is logged for compliance and process improvement.

​
What operator review is

Operator Review is the human-in-the-loop stage of the AgentFlow pipeline. It exists between AI scoring and billing or handoff, and nothing downstream can proceed until an operator makes an explicit decision on each lead. This design is intentional. AI scoring produces a structured, consistent signal — but signals are not decisions. Business context, strategic priorities, edge cases, and relationship history all belong to the operator. The review stage is where your team applies judgment to the AI’s output and produces an accountable result.

​
The protected dashboard

The operator dashboard requires authentication. Only team members with a valid operator account can access lead records, qualification scores, or decision tools. Authentication is enforced server-side; there is no public or guest view of the review queue.
When an authenticated operator opens the dashboard, they see a review queue containing all leads that have completed AI qualification and are awaiting a decision. Each entry in the queue shows:
  • Lead data — the original form fields submitted by the lead, presented in a readable format
  • AI score — the numeric qualification score (0–100) returned by the AI evaluation
  • Confidence level — the model’s stated certainty in its score (low / medium / high)
  • Fit and risk signals — the specific positive and negative indicators the model identified
  • AI summary — a plain-language explanation of the score reasoning
  • Intake metadata — submission timestamp, entry channel, and referral source where captured
Operators see everything they need to make an informed decision without switching between tools.

​
Decision options

Operators can take one of four actions on any lead in the review queue:

Approve

Mark the lead as qualified and ready to proceed. An approval triggers the next pipeline stage — either a billing checkout flow or direct CRM handoff, depending on your configuration. Approved leads move forward immediately.

Reject

Remove the lead from the active pipeline. A rejection records the decision with a timestamp and optional reasoning note. Rejected leads are retained in the audit log but do not proceed to billing or handoff.

Escalate

Flag the lead for senior review or a specialized team member. Escalated leads remain in the queue with an escalation status visible to other operators. Use escalation for high-value or ambiguous leads that require a second opinion before a final decision is made.

Request More Info

Hold the lead and trigger a follow-up request. Use this option when the submission lacks sufficient detail to make a confident decision. The lead stays in a pending state until the requested information is provided and the operator reviews the updated record.
Each decision is a single, deliberate action — operators cannot partially approve or apply multiple statuses simultaneously. This keeps the audit trail clean and unambiguous.

​
The audit trail

Every decision made in the operator dashboard is written to an immutable audit log. Each log entry contains:
  • The lead identifier
  • The decision made (approve / reject / escalate / request more info)
  • The operator who made the decision
  • The timestamp of the decision
  • An optional reasoning note added by the operator
The audit trail cannot be edited or deleted. This is a deliberate design constraint — the log exists to provide a durable, trustworthy record of how your team handled every lead.

​
Why the audit trail matters

The audit log serves two distinct purposes: Compliance — if your business operates under procurement policies, data-handling regulations, or internal governance requirements, the audit trail gives you a timestamped record of every human decision made on every lead. You can demonstrate that AI scoring produced a recommendation and that a human made the final call. Process refinement — reviewing historical decisions alongside AI scores reveals patterns. If operators consistently override high-confidence approvals from the AI, that is a signal your qualification criteria need adjustment. If escalations cluster around a particular lead type, that is a signal for a new decision rule. The audit log is the primary data source for improving your qualification workflow over time.

​
How to use the review queue efficiently

A few practices that keep the review queue moving:
  • Act on high-confidence scores first. Leads where the AI returned a high-confidence score with a clear approval or rejection signal are the fastest to process. Review those first and reserve more time for low-confidence or escalated records.
  • Add reasoning notes consistently. Even brief notes on rejections and escalations make your audit log significantly more useful when you review it later.
  • Set escalation conventions. Agree with your team on what types of leads trigger escalation versus outright rejection. Consistent escalation use makes the queue easier to manage and the audit log easier to interpret.
  • Review the queue on a regular cadence. Leads waiting in the review queue are not progressing to billing or handoff. The faster your team processes the queue, the tighter your lead-to-revenue cycle.

​
Next steps

To understand how the review stage fits into the broader workflow, see The Five-Stage Workflow Pipeline.