QualifyAI agents improve over time through feedback be firm users. Firms compare expected results with predicted outcomes, review the AI’s reasoning, and provide guidance when something looks off. QualifyAI then proposes criteria refinements, which are saved as new versions so firms can track improvements and maintain control.
Why Ongoing Maintenance Matters
QualifyAI is designed to improve through guided feedback, not operate as a static rule set. Regular review ensures evaluations stay aligned with your firm’s evolving priorities and intake decisions.
Criteria Validation & Validation Pool
Validation allows firms to confirm that an agent evaluates leads as expected.
During validation, firms:
Select specific matters for your validation pool to evaluate against the current criteria
Compare expected outcomes with predicted categories
Review reasoning when results don’t match expectations
Provide specific feedback on why the lead should have been evaluated differently
Validation can be done using Test matters with known outcomes or with real leads already in the system.
Providing Feedback to QualifyAI
Feedback tells QualifyAI why an evaluation was correct or incorrect.
Firms can provide feedback when:
A lead is categorized incorrectly
Confidence seems too high or too low
Reasoning misses an important factor
Clear, specific feedback helps QualifyAI propose more accurate rule adjustments.
Refining Criteria
Based on feedback, QualifyAI may:
Suggest updates to existing criteria
Adjust how factors are weighted
Clarify category boundaries
Firms review these proposed changes before saving them.
Agent Versions
Each time refinements are saved, QualifyAI creates a new agent version.
Versions allow firms to:
Track how criteria evolve over time
Compare performance between versions
Maintain a clear history of changes
One version is always set as the default. The default version is recommended as the selected version in automations.