System Overview
This system automatically validates consumer leads every minute using a multi-layered approach. It processes leads that are 5 to 120 minutes old, working in small batches to ensure efficiency and prevent system overload. The goal is to move qualified leads from 'Initial' to 'Pending Allocation' without human intervention.
Operational Efficiency
Reduces manual QA workload by up to 80% and applies consistent quality standards 24/7.
Quality Assurance
Uses regex, phone verification, and ML classification for robust, multi-layer validation.
Scalability
Queue-based processing and configurable batch sizes allow the system to handle volume spikes with ease.
The Automated Processing Workflow
Industry & Basic Checks
The process begins by verifying that QA automation is enabled for the lead's industry and that all baseline consumer data (name, email, etc.) is present and valid.
➡️
➡️
Phone Classification & Regex
The lead is routed based on its phone status (verified vs. unverified). Then, a comprehensive set of regex rules are applied to filter out profanity, test data, and invalid formats in names and emails.
Industry-Specific Processing
For specialized industries like Solar, the system employs machine learning models to classify leads based on criteria like shade coverage and home type, using confidence thresholds to ensure accuracy.
✅ / ❌
Approval or Review
If a lead passes all checks, it's auto-approved, marked as "Good to Sell," and its status is changed to 'pending_allocation'. If it fails any step, it's automatically moved to 'pending_review' for manual QA.
Validation Rules Example: Name Field
The system uses a series of regex patterns to validate data. Here's an example of the logic applied to a consumer's first or last name.