Documents

InvoiceParser AI

Invoice PDFs in, validated structured data out — ready for Excel, PostgreSQL or an ERP.

The problem

Accounting teams re-type the same fields from every invoice: customer, line items, taxes, totals, withholdings. It is slow, repetitive and easy to get wrong.

How it works

A LangGraph pipeline validates each PDF, picks the most reliable text extractor, normalizes the text and asks an LLM for a strict JSON schema. A FastAPI backend processes jobs in the background, and a lightweight interface lets people review and correct fields before exporting.

Pipeline

01Ingestion
02Text extraction
03Cleaning
04LLM → JSON
05Review & export

Highlights

Automatic fallback between three PDF extraction libraries
Tokens, cost and processing time tracked per document
Editable review form before saving to Excel or PostgreSQL
Background jobs with status polling; Dockerized multi-stage build

Stack

LangGraphOpenAIFastAPIPydanticPostgreSQLDocker

Want something like this for your company?

I build production versions of these systems on your data and your infrastructure.

Related service: Document processing

Book a call