How srujangowda5/finance-assistant2 works
Project type and purpose
• This is a full-stack, modular, microservices-based financial assistant.
• Purpose: deliver daily Asia tech market briefings via a voice-enabled Streamlit frontend, powered by multiple FastAPI “agents.”
File structure
• Top-level directories:
• agents/ (each subfolder is a FastAPI microservice)
• orchestrator/ (router that ties agents together)
• streamlit_app/ (frontend UI)
• data_ingestion/ (ETL pipelines for raw data)
• docs/, requirements.txt, Docker config files
• Each agent folder contains its own main.py, helper modules, and possibly data.
• The orchestrator routes requests to the correct agent endpoint.
• streamlit_app drives the UX and makes HTTP calls to orchestrator.
• External dependencies and configs are in requirements.txt, .devcontainer, Dockerfile.
README insights
• Architecture overview table explicitly lists each agent and its role.
• Technologies: FastAPI, Streamlit, OpenAI GPT-3.5, FAISS, yFinance, gTTS, Whisper.
• Data flows: user → Streamlit → orchestrator → agents → external APIs/DBs → orchestrator → Streamlit → TTS.