A local AI that types up a Brazilian accountant's receipts (and knows when it's wrong)

说明与用途

一套完全本地化的 AI 系统每周处理约 100 份巴西会计文件，字段准确率达 95%，将人工录入时间缩短至约 5 分钟。

问题

A self‑employed accountant in Brazil receives ~100 documents a week from small‑business clients and currently types each one into a spreadsheet by hand.

结果

The system achieves 95% field accuracy on 10 fictitious documents, with all problematic documents routed to review. Average processing time is 2.2 s per text PDF and 4.0 s per image, totaling about 5 minutes for a week’s worth of documents.

工作流

• Drop files into a Streamlit page and select the accounting month.
• Each document is read by an open‑weight model running locally in Ollama.
• Python validates the model output and flags rows that need human review.
• The accountant edits flagged rows in an editable table and exports the result to Excel (entries sheet and summary by category).

证据

A self‑employed accountant in Brazil gets ~100 documents a week from her small‑business clients: NF‑e invoices (PDF), NFC‑e receipts photographed on a phone, Pix transfer screenshots, and the occasional handwritten recibo.

来源

Each document is read by an open‑weight model running locally in Ollama.

来源

Python validates everything the model returned and flags the rows a human must check.

来源

95% of fields correct. All three misses come from the blurry receipt.

来源

Average time on an AMD RX 9060 XT (Ollama, Vulkan): 2.2 s per text PDF, 4.0 s per image. That's about 5 minutes for a week of documents.

来源