Core product
Qualification Flows
This page documents how qualification works in the current OrcaPulse project: prompt-driven discovery, intent capture, review logic, lead success evaluation, call scoring, and the operator surfaces used to inspect outcomes.
What qualification means in this project
Qualification in OrcaPulse is the logic used to decide whether an inbound lead is worth advancing, what information is still missing, and which workflow action should happen next.
The current project already expresses qualification in multiple places: workflow and quickstart prompts, social lead AI analysis, lead success evaluation, call scoring, and operator review screens like `Review Lead` and `Lead Hub`.
Core qualification signals
Qualification flows in this project are built from concrete signals, not only from a single yes or no flag. Social leads can carry AI-derived intent, sentiment, and follow-up need. Main leads and calls can carry success evaluation and call scoring.
This gives operators and workflows multiple ways to decide whether a lead is qualified, incomplete, low quality, or ready for the next action.
Social conversations
The project captures from comments, DMs, story replies, story mentions, and platform-specific message surfaces depending on the connected integration.
LinkedIn Lead Gen Forms
LinkedIn uses a dedicated Lead Gen Forms flow with ad account selection, form selection, and sync behavior stored on the integration record.
Main lead records
Once a social lead is accepted and assigned, it is converted into the main Lead model so workflows, timeline events, and execution state can take over.
Prompt-driven qualification
The existing docs flow already treats qualification as prompt-driven discovery. Teams are expected to define the key questions that capture intent, fit, urgency, budget, timeline, location, or use case before launch.
In practice, the first qualification flow should stay short and explicit. The goal is to ask only decision-making questions and connect the answers to the next step, not to turn the interaction into a long survey.
Capture toggles
Each social integration stores booleans for comments, DMs, lead ads, story replies, and story mentions so teams can decide exactly what counts as capture.
Auto-reply and AI
Auto-reply, trigger keywords, AI extraction, AI reply, and AI DM conversation settings are all first-class configuration points in the integration model and UI.
Duplicate protection
Duplicate prevention can be enabled on integrations and workflows so repeated contact data does not keep creating new leads or re-running the same path.
- Ask only decision-making questions: Focus on intent, fit, urgency, budget, timeline, location, or use case.
- Keep prompts natural: Qualification should read like real follow-up messages rather than a rigid form.
- Define thresholds early: Decide what counts as qualified, unqualified, incomplete, or review-required before launch.
- Map answers to actions: Every important answer should connect to a routing, sync, messaging, or human-review outcome.
Review and scoring
Qualification continues after the first conversation. Calls can be marked with `successEvaluation` and scored from 1-10 based on transcript quality and goal achievement. The scoring service uses transcript analysis plus summary context to judge how well the conversation met its goals.
This means qualification is not only what the lead said first. It also includes whether the AI or operator actually moved the lead toward the intended outcome.
Operator visibility
Operators already have review surfaces that support qualification work. `Review Lead` shows lead status, execution state, recall state, call outcome, and success evaluation. `Lead Hub` gives a broader list view across workflows.
Timeline events on the main lead record make qualification observable after the conversation ends. They help teams inspect what was sent, what succeeded, what failed, and how the lead moved through the workflow.
- Review Lead: best for detailed inspection of one lead’s qualification and follow-up state.
- Lead Hub: best for seeing patterns across many leads and workflows.
- Timeline history: best for understanding how qualification translated into actual actions.
How to design the first flow
A strong first qualification flow in this project usually follows a simple pattern: capture inbound intent, collect the minimum missing business details, decide whether the lead is qualified or incomplete, and then route the result into the right next action.
Most teams should start narrow. One short path with clear qualification thresholds is easier to test and improve than a large branching system.
- Start with one qualification path: avoid too many branches in version one.
- Use explicit outcomes: qualified, unqualified, incomplete, or review-required should be easy to tell apart.
- Route immediately after decision: connect qualification output to CRM sync, webhook sync, messaging, AI call, or operator review.
Next steps
After qualification logic is clear, the next layer is routing and follow-up. That means deciding which qualified leads should sync to a CRM, trigger a webhook, receive messaging, start AI calling, or wait for human review.
Keep the first qualification flow small, verify the review surfaces show the right signals, and then continue into routing rules and follow-up automation.