shbshahriar/sequential-knowledge-synthesizer repository overview

Sequential Knowledge Synthesizer is a lightweight AI-powered web app that breaks down any topic into three structured outputs — an outline, detailed notes, and a final summary — using a sequential LangGraph workflow backed by Google's Gemini LLM.

How it works:

1. You enter a topic in the browser UI.
2. The LangGraph workflow runs three nodes in order:
   • generate_outline → builds a structured outline of the topic
   • generate_notes → expands each outline point into detailed notes
   • generate_summary → condenses everything into a concise summary
3. All three outputs are streamed back and displayed in the UI.

Tech stack:

• Backend : FastAPI, LangGraph, LangChain, langchain-google-genai
• LLM : Google Gemini (gemini-3.1-flash-lite-preview)
• Frontend: Single-file HTML/JS served by FastAPI
• Packaging: uv, multi-stage Docker build (python:3.13-slim)

Quick start: docker run --env-file .env -p 8000:8000 shbshahriar/sequential-knowledge-synthesizer

Then open http://localhost:8000⁠ in your browser.