Knowledge Roadmap Builder is a Python + Streamlit + agentic AI app that generates course/learning roadmaps as interactive graphs based on a simple natural-language prompt from the user.
- Takes a user prompt like:
"I want a roadmap to learn full‑stack web development" - Uses an AI agent to:
- Break the goal into topics, subtopics, and learning stages
- Order them into a sensible sequence (prerequisites → core → advanced)
- Add short descriptions and suggested resources (optional)
- Renders the result as a graph/roadmap directly in the Streamlit UI.
- Python 3.11+
- Streamlit for the web UI
- Agentic AI stack (e.g. Langgraph) to orchestrate roadmap generation
- A hosted LLM API (Euron API key(OpenAI))
- For google search serpapi and wikiapi
- Graph/visualization library (
graphviz)
- Natural-language input: “Data Science roadmap”, “DSA roadmap for interviews”, “NLP roadmap”, etc.
- AI-generated nodes:
- Milestones (Beginner, Intermediate, Advanced)
- Topics and subtopics
- Dependencies between concepts
- Interactive graph view:
- Zoom / pan
- Click nodes to see details (description, notes, resources)
- One-click regenerate / refine:
- “Add more math topics”
- “Make it 6-month plan”
- “Focus on Python only”
-
User input
User describes the roadmap they want (domain, level, constraints). -
Agent planning
An AI agent:- Analyzes the goal
- Plans stages (Foundations → Core → Projects → Advanced)
- Expands each stage into nodes and edges
-
Graph construction
Nodes and edges are converted into a graph data structure. -
Visualization
Streamlit renders the graph and any metadata (tables, sidebars, descriptions).
git clone https://github.com/<your-username>/knowledge-roadmap-builder.git
cd knowledge-roadmap-builder
python -m venv .venv
source .venv/bin/activate # On Windows: .venv\Scripts\activate
pip install -r requirements.txt
streamlit run app.py