I built Thelawin because EU e-invoicing is one of those boring problems that gets surprisingly deep once you ship it.
The basic idea is simple: send invoice JSON, get an EU e-invoice back.
It currently supports ZUGFeRD, Factur-X, XRechnung, UBL, CII, Peppol, FatturaPA and plain PDF output. For PDF-based formats it returns PDF/A-3 with embedded XML. There is also a validation endpoint, so you can check the JSON before generating the invoice.
The API is stateless by design. Invoice data is processed, returned, and not stored. Usage logs do not contain invoice contents.
I also added an MCP endpoint. The goal is that an agent can pass sender, buyer and invoice-line data to a schema-bound tool instead of trying to produce invoice XML itself.
I would be interested in feedback from people who have dealt with e-invoicing, ERP integrations, accounting software, Peppol, or PDF/A-3 generation. The hard part is not making a PDF. The hard part is avoiding all the small format mistakes that only show up when a real customer tries to submit the invoice somewhere.
0 comments
—
Show HN: Thelawin – Generate EU e-invoices from JSON · HackerTrans
I built Thelawin because EU e-invoicing is one of those boring problems that gets surprisingly deep once you ship it.
The basic idea is simple: send invoice JSON, get an EU e-invoice back.
It currently supports ZUGFeRD, Factur-X, XRechnung, UBL, CII, Peppol, FatturaPA and plain PDF output. For PDF-based formats it returns PDF/A-3 with embedded XML. There is also a validation endpoint, so you can check the JSON before generating the invoice.
The API is stateless by design. Invoice data is processed, returned, and not stored. Usage logs do not contain invoice contents.
I also added an MCP endpoint. The goal is that an agent can pass sender, buyer and invoice-line data to a schema-bound tool instead of trying to produce invoice XML itself.
There is a public sandbox flow here: https://thelawin.dev/ai
I would be interested in feedback from people who have dealt with e-invoicing, ERP integrations, accounting software, Peppol, or PDF/A-3 generation. The hard part is not making a PDF. The hard part is avoiding all the small format mistakes that only show up when a real customer tries to submit the invoice somewhere.