Claude AI payroll automation system builder (Singapore CPF/MOM, Dropbox, Python)

We are a Singapore company (F&B / retail, 100+ employees) looking for an expert to build an AI-powered monthly payroll automation system. Today payroll takes ~7 days of manual work; the goal is to bring it under 1 working day with high accuracy and automated error detection. The system must scale toward 500+ employees. WHAT WE NEED YOU TO BUILD - Automated file intake from Dropbox (attendance, schedules, commission, incentives, employee master, CPF folders) - Intelligent extraction from Excel, CSV, PDF, and image/screenshot schedules (OCR) - Payroll calculation engine: basic pay, overtime (Singapore MOM rules), commission, performance incentives, and deductions (unpaid leave, advances, penalties) - CPF calculation (employee + employer, age-based rates, ceilings) and CPF EZPay-compatible export file - Automated error/anomaly detection (missing attendance, salary/CPF mismatch, negative pay, excessive OT, missing bank/NRIC/FIN, duplicate payouts) - Auto-generated PDF payslips (password-protected) plus payroll summary, CPF summary, and audit-trail reports REQUIRED SKILLS / EXPERIENCE - Anthropic Claude API and/or other LLM-based document extraction workflows - Python automation, Dropbox API integration - OCR tooling - Prior payroll / HR-tech work strongly preferred; Singapore CPF & MOM compliance knowledge is a big plus HOW WE'D LIKE TO WORK - Please start with a small paid test task (e.g. extract and calculate one month for a sample of employees) before the full build - Clear documentation and an audit log for every calculation are essential; original payroll files must never be overwritten - Good communication and a proposed step-by-step implementation plan in your proposal Please share relevant past projects (payroll, document-extraction, or Claude/LLM automation) when you apply.