This project implements an autonomous AI agent capable of:
- Understanding natural language requests
- Creating its own execution plan
- Executing the plan
- Performing a lightweight self-review
- Generating a professional Microsoft Word document
- Python
- FastAPI
- Groq Llama 3.3 70B
- python-docx
- Pydantic
app/
main.py
agent.py
planner.py
executor.py
reflection.py
doc_generator.py
llm.py
config.py
models.py
prompts.py
pip install -r requirements.txt
Create .env
GROQ_API_KEY=YOUR_KEY
Run
uvicorn app.main:app --reload
Open
{ "request":"Create a business proposal for an AI Healthcare Startup." }
Implemented lightweight Reflection/Self-check.
The generated document is reviewed before returning the final result.
-
Tool Calling
-
RAG
-
Conversation Memory
-
Retry based Reflection
-
Multi-Agent Architecture
uvicorn app.main:app --reload streamlit run app/streamlit_app.py