AI Payroll Exception Assistant
Payroll anomaly detection, with LLM explanations planned
Payroll exception review tool in development — the data model and anomaly detection rules are built; plain-language LLM explanations are the next phase.
Problem
Payroll teams manually review hundreds of exception reports each cycle, burning hours on pattern recognition that could be automated.
Goal
Create a tool that ingests payroll data, flags statistical outliers, and provides LLM-generated explanations so reviewers can act faster with higher confidence.
What I Built
A Next.js application with a structured data pipeline that feeds payroll records through anomaly detection rules — designed to surface flagged items with OpenAI-generated plain-language summaries once the LLM phase is wired in.
Key Features
- Automated exception detection across overtime, missed punches, and pay variances
- Planned: LLM-generated plain-language explanations per flagged item
- Planned: Batch review workflow with approve/escalate/note actions
- Planned: Audit trail for compliance and manager sign-off
- Planned: Dashboard summary with trend charts
Challenges
- Designing PII-safe prompts so sensitive payroll data never reaches external APIs
- Tuning anomaly thresholds to minimize false positives without missing real issues
- Planning LLM prompts that produce consistent, auditable explanations
Outcome
Data model and anomaly detection rules are built and tested. Next phase wires in LLM-powered explanations with PII-safe prompt design.
Current Status
Next phase: LLM explanations with PII-safe prompts, then the review workflow and audit trail.