# LLM translation Projects at AI Tinkerers

> Canonical HTML: https://aitinkerers.org/technologies/llm-translation
> Markdown URL: https://aitinkerers.org/technologies/llm-translation.md
> Technology record last updated: 2026-08-12T04:36:24Z
> Generated: 2026-09-22T05:46:10Z

LLMs are revolutionizing translation, moving beyond traditional NMT to capture nuanced meaning and context.

Large Language Models (LLMs) are transforming machine translation by offering a more sophisticated approach than previous Neural Machine Translation (NMT) systems. Unlike NMT, which often translates sentence-by-sentence, LLMs leverage their extensive training data and contextual understanding to produce more fluent and accurate translations, grasping subtleties like idiom and tone. For example, Google Translate (which now incorporates LLM advancements) can better differentiate between homonyms or translate culturally specific phrases with greater precision. This shift enables higher-quality localization and more natural cross-lingual communication, impacting everything from customer service chatbots to global business operations.

- Official technology site: https://www.deeplearning.ai/the-batch/llms-for-translation/
- Public AI Tinkerers demos and talks: 1
- Result page: 1 of 1

## Recent Public Talks and Demos

### [Translating a book is a document problem, not a prompt problem](https://da-nang.aitinkerers.org/talks/rsvp_DOjIyS5U2Z8)

Translating a paragraph with an LLM is easy. Translating a real book is a document-understanding problem. Nga shares lessons from building Omniglot, an AI system for translating books and complex documents end to end. Omniglot processes structured PDFs, scans and images with challenging layouts, including writing systems with vertical and right-to-left reading order, before reconstructing documents into a representation suitable for translation. She will cover how the system combines document parsing, contextual translation memory, editor preferences and an embedded AI assistant that helps translators investigate terminology and compare translation choices. The talk explores a broader lesson from building applied AI products: the hardest problems often aren't the model call itself, but perception, context, memory, and designing AI that helps humans make better decisions. Nga Nguyen is the founder of Consen.AI and the builder of Omniglot. She works at the intersection of translation, machine learning and human-AI interaction, and is also a published Vietnamese translator. Omniglot was a Global Semifinalist at Microsoft Imagine Cup 2025.

- Event context: AI Tinkerers Da Nang: Inaugural Meetup — 2026-08-15 — Da Nang
- Public talk page: https://da-nang.aitinkerers.org/talks/rsvp_DOjIyS5U2Z8

## Related Technologies

- [Document parsing](https://aitinkerers.org/technologies/document-parsing) ([Markdown](https://aitinkerers.org/technologies/document-parsing.md)) — 1 public demo
- [OCR](https://aitinkerers.org/technologies/ocr) ([Markdown](https://aitinkerers.org/technologies/ocr.md)) — 10 public demos
- [Retrieval](https://aitinkerers.org/technologies/retrieval) ([Markdown](https://aitinkerers.org/technologies/retrieval.md)) — 4 public demos
- [Translation memory](https://aitinkerers.org/technologies/translation-memory) ([Markdown](https://aitinkerers.org/technologies/translation-memory.md)) — 1 public demo