Home

Jewish Learning Suite

AI-powered tools for ancient texts

Making classical Jewish scholarship accessible through technology. Real-time translation, linguistic analysis, and AI-assisted study.

לעילוי נשמת

אברהם חיים בן דוד

L'ilui Nishmas Avraham Chaim ben David

The Projects

Talmudic Study App

Real-time AI streaming translation

GitHub →

Watch scholarly translations appear word-by-word as Gemini processes Hebrew/Aramaic source texts. Inline commentary (Mefarshim) with nested exploration. Study journey tracking.

Speed

200 chars/sec

Model

Gemini 2.5 Flash

Layout

70/30 EN/HE

Ohr Avraham Chaim

Professional Hebrew/English book typesetting

GitHub →

LaTeX + AI translation. Professional bilingual editions of classical texts. Proper RTL formatting. Print-ready PDF output.

Projects:

• Tur & Mefarshim (~50 simanim)
• Nefesh HaChaim (Gate I complete)
• Derech Hashem (in progress)
• Siddur Analysis (858 words, 248 roots)

MCP Sefaria Server

Torah texts through AI

GitHub →

MCP server for querying the Sefaria Jewish text library. Access any text in the Jewish canon through Claude or any LLM.

Talmudic Study App Architecture

Sefaria API → Hebrew/Aramaic Text → Gemini 2.5 Flash → Streaming → React UI
     ↓              ↓                       ↓                  ↓
Commentary    Section Parser        Real-time Chunks    Progressive Display
  Links        (numbered)          (200 chars/sec)      (green highlights)

Key Features:

• Server-Sent Events for streaming
• Translation persistence (Supabase)
• One commentary per section rule
• Nested commentary depth tracking
• User journey analytics
• Rate limiting (Upstash Redis)

Siddur Analysis System

Comprehensive Linguistic Analysis

858

Words analyzed

248

Unique roots

41

JSON files

6.2MB

Structured data

12 fully analyzed prayers (Ashrei, Aleinu, Shema, morning blessings, Torah blessings). 36 Python scripts for linguistic analysis. 6 English translation sources compared. Pure linguistic processing—no AI.

Tech Stack

Talmudic App

Next.js 15, React 19, TypeScript, Supabase, Gemini via OpenRouter, Tailwind 4, shadcn/ui

Ohr Avraham Chaim

XeLaTeX, Python, Sefaria API, OpenRouter/Gemini, babel/polyglossia for Hebrew

Siddur Analysis

Python, JSON, shoresh (root) mapping, cross-text frequency tracking

MCP Sefaria

MCP Protocol, Sefaria API, Node.js

Use Cases

Yeshiva Students

Study Talmud with instant English translation

Adult Learners

Access classical texts without Hebrew mastery

Rabbis & Educators

Prepare lessons with AI translation support

Global Community

Learn Talmud regardless of language background

The Vision

"Making classical Jewish scholarship accessible through technology. Not replacing the traditional learning experience—extending it to those who couldn't access it before."

Translation Pipeline

  1. Fetch — Download Hebrew text from Sefaria API
  2. Translate — Use Gemini 2.5 Flash Lite via OpenRouter
  3. Generate — Create LaTeX files with proper formatting
  4. Compile — Build PDFs with XeLaTeX

Prompt: "Output ONLY the translation" — no intro phrases, no explanations.

Talmudic Study App → · Ohr Avraham Chaim → · MCP Sefaria →