Accounts Payable Automation

Part of the worlds-biggest-software-project initiative.

An open-source, AI-native accounts payable platform that combines invoice capture, three-way matching, approval workflows, duplicate detection, and payment scheduling in a single transparent stack.

Accounts Payable Automation targets finance teams stuck between expensive enterprise P2P suites and underpowered open-source ERP modules. It is built for AP managers, controllers, and CFOs at SMB and mid-market companies whose invoice volumes are outpacing headcount but who cannot justify $100K+ enterprise vendor spend. The project unifies the capture, coding, approval, and payment chain that today requires stitching together multiple closed-source vendors.

Why Accounts Payable Automation?

The AP automation market is $6.94B in 2026, growing at 12.44% CAGR to $12.46B by 2031, yet only 5% of AP teams are fully automated and 74% are only partially automated (ACARP 2024) — a structural greenfield.
Manual processing costs $12–$20 per invoice and takes 10–30 minutes; AI automation cuts cost to ~$2.36 per invoice and processing time to 1–2 seconds, but those gains are gated behind closed-source vendors.
Existing open-source ERPs (ERPNext, Odoo Community) provide AP primitives but ship with no invoice OCR, no adaptive GL coding, and no payment network — leaving teams to integrate proprietary tools or stay manual.
Commercial leaders lock customers in: BILL and AvidXchange via proprietary payment networks (250,000+ enrolled vendors each), Stampli via opaque "Billy the Bot" coding models that cannot be inspected or exported.
EU, Australia, and Singapore are mandating PEPPOL e-invoicing, and no GPL-licensed AP tool natively supports PEPPOL today — only Odoo Enterprise does, behind a commercial licence.

Key Features

Invoice Capture and Data Extraction

AI-powered OCR for header and line-item extraction from PDF, image, and email-attachment inputs
Handles non-standard and irregular invoice formats using multi-modal vision-and-text models
Confidence scoring on extracted fields with low-confidence items routed for human review
Document classification across invoice types and languages

Matching, Coding, and Workflow

Three-way matching engine: purchase order, goods receipt, and supplier invoice
Adaptive AI GL coding that learns company-specific patterns from approval history
Configurable multi-level approval workflows with delegation, escalation, and conditional routing
Invoice-level collaboration: comments and approver queries attached to the invoice image
Email and mobile notifications for pending approvals

Duplicate, Anomaly, and Fraud Detection

Exact and near-duplicate detection across invoice number, amount, vendor, and dates
Semantic similarity scoring on OCR-extracted text to catch sophisticated duplicate billing
Behavioural anomaly scoring on vendor payment patterns to flag ghost vendors and billing fraud
Full audit trail covering the complete invoice lifecycle

Vendor Management and Payments

Vendor onboarding with W-9 / 1099 tax form tracking
Multi-currency AP with exchange rate management
Automatic payment scheduling based on invoice due dates
Payment execution integration via Open Banking APIs (ACH, bank transfer)
Vendor self-service portal for invoice submission and payment status inquiry

Compliance and ERP Integration

PEPPOL / UBL e-invoicing for EU, Australian, and Singaporean mandates
ERP synchronisation with GL coding push-back (ERPNext, NetSuite, QuickBooks connectors planned)
REST API for all documents and transactions
Optional transparent ML model export so customers can audit and own their coding model

AI-Native Advantage

Existing OCR vendors (ABBYY, Rossum) handle capture, and workflow vendors (Stampli, BILL) handle approvals, but no project integrates the full chain — capture, PO matching, GL coding, approval routing, duplicate detection, and payment scheduling — under a single AI agent layer. Adaptive GL coding here is inspectable and exportable, unlike the opaque vendor-locked models used by Stampli's Billy the Bot or Tipalti's AI Assistant, which is critical for audit defensibility. Duplicate and fraud detection apply semantic similarity across OCR text, vendor behavioural profiles, and payment timing patterns, catching cases that exact-match rules miss.

Tech Stack & Deployment

The project is designed for self-hosted deployment with optional managed hosting, drawing on the ERPNext and Frappe model where all features ship in the open-source core with no premium tier. Integrations target the standards already required by the category: PEPPOL BIS Billing 3.0, UBL (ISO/IEC 19845), EDI X12 810 for US retail and manufacturing, and ISO 20022 for bank payment messaging. ERP connectivity is API-first via REST, with planned native connectors for ERPNext, NetSuite, and QuickBooks.

Market Context

The AP automation market is $6.94B in 2026 and projected to reach $12.46B by 2031 at 12.44% CAGR (Mordor Intelligence). SAP, Oracle, Coupa, Tipalti, and Basware collectively hold roughly 30% share, with enterprise pricing in the $100K+ range and mid-market vendors largely refusing to publish list prices. Primary buyers are AP managers and controllers at $50M–$500M revenue companies, SMB CFOs seeking cost-per-invoice reduction, and finance IT evaluators focused on ERP integration depth.

Project Status

This project is in the research and specification phase.
Contributions, feedback, and domain expertise are welcome.

Contributing

We welcome contributions from developers, domain experts, and potential users. See CONTRIBUTING.md for guidelines.

Important: All contributions must be your own original work or clearly attributed open-source material with a compatible licence. Copyright infringement and licence violations will not be tolerated and will result in immediate removal of the offending contribution. If you are unsure whether a piece of code, text, or other material is safe to contribute, open an issue and ask before submitting.

Licence

Licence to be determined. See discussion for context.