AI-Powered Expense Validation & Auditing

An enterprise-scale AI automation system for Oracle Expense Report validation and auditing.

I architected the full solution: model selection, event-driven integration topology, and the orchestration layer that extracts receipt data, validates against policy rules, and generates audit reports at scale.

The client's finance team was spending significant time manually reviewing and auditing expense reports. The process was slow, inconsistent, and couldn't scale with the volume of submissions.

Validating receipts against expense claims and ensuring policy compliance required tedious manual comparison that was prone to human error.

I architected an event-driven AI system on OCI. I evaluated multiple models through OCI Generative AI and selected Llama 4 Maverick for its multi-document processing capabilities. The architecture integrates Oracle VBCS for configurable audit rules, Oracle Integration Cloud for event-driven triggers, and a Python backend on OCI Compute that orchestrates extraction, validation, and report generation workflows.

The system handles multiple receipts per attachment, validates expenses against configured rules, and attaches audit reports directly to expense records in Oracle Fusion.