How viswadarshan-024/finance-agentic-ai works
Project Type & Purpose
• This is a full-stack-ish financial research application developed in Python. The tool retrieves real-time stock data (via yfinance), performs AI-driven financial analysis (using Groq’s language model), augments insights with web search data (via Google Custom Search API), and then presents interactive visualizations (using Plotly) to end users through a Streamlit-based UI.
• The primary goal is to empower investors with quick, comprehensive, and intelligent stock market insights.
Main Components
• Frontend/UI: The Streamlit user interface that drives interactivity, displays charts, and shows AI-generated markdown reports.
• Stock Data Retrieval Module: Uses the yfinance package to fetch real-time stock information and financial metrics.
• AI Analysis Module: Integrates with Groq’s language model (Llama 3.3 70B) to generate advanced financial analysis.
• Web Search Integration Module: Leverages the Google Custom Search API to pull supplementary research and news.
• Data Visualization Module: Utilizes Plotly to produce dynamic, interactive charts for price trends.
• Security/API Management: Manages and stores API keys securely via Streamlit’s secrets management and implements error handling around external API interactions.
Establish Relationships and Data Flow
• User Input (stock ticker, interaction events) goes to the Frontend.
• The UI triggers the Stock Data Retrieval Module, which calls Yahoo Finance’s APIs.
• Retrieved stock data is then routed to both the AI Analysis and Data Visualization modules.
• Separately, the UI or backend initiates a call to the Web Search Integration module to fetch relevant news or research.
• All outputs (AI-generated insights, visual charts, supplementary research) are sent back to the UI for display.
• Error-handling and security checks when interacting with external APIs.
Architectural Patterns and Key Technologies
• Although implemented in a single Python file (main.py) for demonstration, conceptually separate components can be shown using a layered design: Presentation (Streamlit UI) → Business Logic (data retrieval, analysis, search integration) → External Services (Yahoo Finance, Groq AI, Google Custom Search, Plotly).
• This logical separation hints at a modular design despite the minimal file count.
• That the project uses dependency management via requirements.txt.