# A local AI for Gramps?

**URL:** <https://gramps.discourse.group/t/une-ia-locale-pour-gramps/7187>\
**Category:** Français (French)\
**Tags:** ai, ia\
**Created:** [March 9, 2025, 10:25am UTC](https://gramps.discourse.group/t/une-ia-locale-pour-gramps/7187 "2025-03-09T10:25:34Z")\
**Posts on this page:** 1\
**Showing post:** 21

<div class="post-metadata">

**Author:** ![romjerome](https://yyz2.discourse-cdn.com/free1/user_avatar/gramps.discourse.group/romjerome/32/1489_2.png) [@romjerome](https://gramps.discourse.group/u/romjerome)\
**Post date:** [January 22, 2026, 12:33pm UTC](https://gramps.discourse.group/t/une-ia-locale-pour-gramps/7187/21 "2026-01-22T12:33:49Z")

</div>

I’m more or less doing the same kind of tests as

> [@Gramps plugin concept: OCR + name index for large document collections](https://gramps.discourse.group/t/gramps-plugin-concept-ocr-name-index-for-large-document-collections/9031):
>
> I have a database with about 14,000 individuals in Gramps. I also have an archive of around 1,700 scanned documents (obituaries, PDFs, images), of which only part are already included in Gramps as sources. I do not intend to import this entire collection wholesale into Gramps. My idea is to develop a Gramps plugin that: scans an external folder of documents, applies OCR and name detection, stores the results in a JSON index (document → names found), and, when a person is selected …

But surname and place name matching remains an issue for these AI systems.

Recently, I knew one of my branches originated from Bavaria (_Bavarians might complain since the territorial boundaries aren’t very clear…_). The surname in French (_well, Frenchified_) is `Haentzler`, sometimes `Häntzler` for local pronunciation! But in Bavaria, it was more likely `Hänßeler`, or even `Hensler`. Apart from the accents on certain letters and local dialect variations, it sounds very similar. Explaining that to a machine? No thanks!

---

_[View the full topic](https://gramps.discourse.group/t/une-ia-locale-pour-gramps/7187)._