Igniters-week5 - ESV Bible companion RAG AI

[nsikanikpoh]

I used langchain, gpt-4.1-nano . Gradio to build an ESV Bible companion that balances word-for-word accuracy with modern, readable English like the ESV Bible does. Handy for a Sunday School teacher, Bible scholar, or personal studies.

• Given a Sunday school topic, the companion generates bible verses that are relevant to the topic.
• Given an excerpt from the Bible, the companion generates bible verses that are relevant to the excerpt
• At the end of every answer, the companion generates what The bible wants us to do as lessons from the bible verse found. @iamumarjaved

[iamumarjaved]

@nsikanikpoh Thank you for the submission. Just for future reference you can improve these few things in the code:

• load_dotenv() is called after embeddings is initialized in ingest.py, move it to the top before any env-dependent code, and remove the unused OpenAIEmbeddings import
• No error handling if KNOWLEDGE_BASE directory doesn't exist or has no .md files will silently produce an empty vector store
• combined_question() concatenates all prior user messages for retrieval, which can pollute queries in long conversations — limit to the last 2–3 turns
• format_context() has a stray print(doc.metadata) debug statement that should be removed before deploying
• History is passed as raw list[dict] into convert_to_messages() explicitly construct HumanMessage/AIMessage objects instead to avoid silent breakage

[iamumarjaved]

@ed-donner This is ready to be merged.