Purchase Order & Invoicing System with ML Automation (PERL)
Complete purchase order and invoicing system with ML-powered automation using AWS and NLP for intelligent line-item matching
The Problem
The Solution
Designed and developed a complete purchase order and invoicing system with modular architecture for future ERP expansion. The system features ML-powered automation using event-driven AWS architecture for email ingestion, document extraction, and intelligent matching between invoices and purchase orders.
Built full-stack system with Laravel backend, PostgreSQL database, and Svelte frontend. Implemented ML-powered automation layer using AWS Lambda, Textract for extraction, and NLP algorithms (tokenisation, TF-IDF, N-gram similarity) for intelligent line-item matching. Designed complete database schema for financial workflows, approval processes, and audit trails.
Tech Stack
Key Features
- Complete purchase order and invoicing workflows
- Modular architecture designed for future ERP expansion
- RESTful API layer
- Event-driven AWS architecture for automation
- Email ingestion with automated processing
- ML-powered document extraction using Textract
- NLP algorithms for intelligent line-item matching (tokenisation, TF-IDF, N-gram similarity)
- Financial workflow management with approval processes
- Comprehensive audit trail system