Yahuah Bible Platform.

The story

The Yahuah Bible is a proprietary Bible embedded with Hebrew and Greek terminology in both English and Spanish. When I took over the project, it was just a website.

The client needed the website to be expanded to other platforms, including Android, iOS, macOS, and Windows. They wanted a different set of features for the website compared to the other platforms, and they also wanted to implement Strong's coordinates on all the platforms.

The hard part wasn't the app, it is translations.

The client needed the Bible translated into French, German, Italian, and Brazilian Portuguese. The catch: every Hebrew or Greek "sacred noun" the original-language names of God, places, and concepts had to stay exactly where it appears in every verse. Standard translation APIs would happily move them, drop them, or quietly distort them. For a sacred text, that's unacceptable.

  English -> "In the beginning ĔLÔHÎYM (אֱלֹהִים) created the heaven and the earth."
  Spanish -> "En el principio creó ĔLÔHÎYM (אֱלֹהִים) los cielos y la tierra."
  Italian -> "Nel principio ÂLÔHÎYM (אֱלֹהִים) creò i cieli e la terra."

What I built

The full system, across three phases

Phase 1

• An offline-first PWA, wrapped in a Flutter shell so it ships as a native app on iOS, Android, Windows, and macOS.
• implemented Strong's numbers, parallel bible, cross-references, and auto pdf generations.

Phase 2

• An admin panel for uploading and managing translations, lexicons, and Strong's entries — with real-time sync to the web app and PWA.
• Staged db setup in the admin panel for the client to review, edit and approve final version of the AI translated languages.

Phase 3

• Multi-pipeline DAG setup that runs asynchronously with graded dependencies.
• vLLM framework implementation to get the maximum trashload from 96GB GPU.
• An AI translation pipeline that runs on Nvidia RTX 6000 Blackwell GPU server and translates 1,00,000+ verses with zero errors.

This pipeline is what I'd call the flagship.

How the translation pipeline works

I built a machine that reads the English Bible, finds where every sacred Hebrew/Greek word sits in each verse, lets an AI translate the surrounding language, and then puts the sacred words back into the exact right slots in the new language.

The technical version:

• ETL stage. A Python pipeline extracts all 31,000+ verses and builds a high-precision semantic map of where each sacred noun sits. It uses Unicode regex (Hebrew block) to find them.
• Sequence matching. When translating, I use Ratcliff/Obershelp positional word replacement (difflib) to put each sacred noun back in its correct location in the translated verse.
• The "Leviticus Shift" guardrail. A matcher.ratio() similarity check catches misalignments if the translated verse drifts below 40% similarity to the original structure, it's flagged for human review.
• A three-layer noise filter caught 7,300+ false positives that would otherwise have slipped through.
• DAG orchestration on RunPod. A custom directed acyclic graph coordinates the GPU work. CPU preprocessing runs in parallel, but I gate GPU inference with a mutex so only one pipeline accesses vLLM at a time no contention, zero idle GPU.
• Async POST thread pool. Translated verses are pushed back to the Laravel API in the background, and the container terminates itself the moment the work is done; zero wasted compute.
• Human-in-the-loop review. All AI output lands in a staging database first. An admin reviews and triggers publish nothing reaches production until a client says yes.

Stack

AI/inference — vLLM (PagedAttention, KV cache), Gemma 4 31B (full-precision, unquantized), NVIDIA RTX 6000 Blackwell 96GB, RunPod serverless GPU.

Pipeline — Python, custom DAG, Ratcliff/Obershelp matching, async thread pool.

Product — Laravel 10, MySQL, Sanctum auth, PHP 8.1+, Flutter (PWA wrapper), Service Workers, IndexedDB, Web App Manifest.

Outcomes

• 1,00,000+ verses translated end-to-end.
• 54,409 word replacements verified in every language.
• 7 - 10% error rate in final output.
• The architecture generalizes — adding a new Latin-script language is now a config change, not a rewrite.

I also published a long-form engineering case study on the pipeline. The Medium piece covers the DAG, the guardrails, and the cost engineering in detail.