Scripture Threads

Scripture Threads maps Bible themes from Genesis to Revelation using AI insights and MongoDB vector search—explore deep patterns through interactive visuals in a fast, responsive web app.

Comment

📘 About the Project

🔥 Inspiration

As someone passionate about both theology and technology, I wanted to create a tool that helps people see the Bible not as isolated verses, but as a unified, living narrative. I was inspired to explore how modern AI could uncover and visualize deep scriptural themes—something that's often missed in traditional study formats. Scripture Threads was born from that vision.

🧠 What I Learned

  • How to integrate Google's Generative AI (Gemini) to generate contextual explanations and theological summaries
  • Building semantic search with MongoDB Vector Search and embedding models
  • Managing a full-stack app solo using React (Vite) for the frontend and FastAPI for the backend
  • Structuring biblical data into a vectorized knowledge base that supports fast, meaningful queries

🛠️ How I Built It

  • Processed and enriched biblical text and commentaries using Google Colab
  • Generated vector embeddings with text-embedding-004
  • Stored structured, enriched data in MongoDB Atlas with native vector search
  • Designed a responsive UI with React + Tailwind CSS, visualizing books and themes
  • Created a FastAPI backend to manage search, content generation, and API responses
  • Deployed the entire system on Google Cloud using Dockerized containers

🚧 Challenges

  • Handling everything solo—from data cleaning to UI/UX to backend logic—was intense but rewarding
  • Designing a flexible schema to represent the rich relationships between scripture and theology
  • Keeping AI usage efficient to stay within API quotas, which I solved using smart caching
  • Creating clear, intuitive visualizations for complex theological ideas with limited resources
Share this project:

Updates