I build OCR and document-processing systems that turn PDFs, scans, emails, invoices, purchase orders and delivery documents into structured, validated data. I've worked on logistics document automation for Delassus / SMATCH, including OCR integration with warehouse workflows, and I built Ironclad OCR / OCRAgent, an open-source intelligent document-processing system designed around extraction, validation, matching, human review and auditability. I can help you with: - Invoice, PO, delivery note and PDF OCR - PDF/Image → JSON, CSV, Excel or API - Structured field and line-item extraction - PO / invoice / delivery document matching - Automatic discrepancy and duplicate detection - Human review for uncertain OCR results - Email → OCR → validation → ERP/WMS workflows - High-volume asynchronous document processing Stack: Python, FastAPI, PaddleOCR, OCR/VLMs, PostgreSQL, Redis, Docker, LangGraph, n8n, REST APIs. If your team is manually reading, typing, comparing or validating documents, I can turn that process into an automated workflow.