How rauhanahmed/financialanalystaiagent works
Project Type & Purpose
• This is a Python-based, full-stack analytical dashboard combining market data retrieval, interactive visualizations, and AI-driven insights.
• The “backend” is a monolithic Streamlit application (main.py) that orchestrates data acquisition, visualization, and AI analysis.
File Structure Analysis
• Top-level scripts
• main.py: Entry point that renders the Streamlit UI and wires up all components.
• financeAnalystAgent.py: Encapsulates the AI agent logic (Google Gemini via Phidata).
• utils.py: Data-fetching (yfinance), chart generation (Plotly), helper functions.
• Configuration
• params.yaml: Model parameters, ticker list URL, agent instructions.
• Environment variables loaded via python-dotenv (.env expected).
• Support files
• requirements.txt, .devcontainer/ for IDE/container setup, demo assets.
README Insights on Architecture
The README already breaks down the data flow:
1. Data Acquisition (Yahoo Finance, CSV ticker list)
2. Data Visualization (Plotly charts)
3. AI-Powered Analysis (Financial Analyst Agent → Google Gemini 2.0 Flash via Phidata)
4. User Interaction (Streamlit dashboard)
5. Configuration (YAML + dotenv)
a. Main Components
• Streamlit UI Layer (frontend)
• Data Layer
• YFinance API
• CSV Ticker List URL
• Visualization Layer (Plotly)
• AI Agent Layer
• FinanceAnalystAgent (Phidata wrapper)
• Google Gemini 2.0 Flash model
• Configuration Layer (params.yaml, .env)
• External Services (Yahoo Finance, Google Gemini endpoint)
b. Relationships & Data Flow
1. Streamlit UI sends ticker request → Utils.fetch_data → Yahoo Finance (yfinance)
2. Retrieved data → Utils.generate_chart → Plotly → Streamlit UI
3. User query in UI → FinanceAnalystAgent → Phidata → Google Gemini → Agent response → UI
4. At startup: main.py loads params.yaml + environment variables → passes config to agent and utils
c. Architectural Patterns & Principles
• Layered approach (UI, business logic, external services)
• Single process “full-stack” app but logically separated into modules
• Configuration-as-code (YAML + dotenv)
• Modular design for easy extension (adding new data sources or analytics tools)
d. Key Technologies & Frameworks
• Streamlit (UI)
• yfinance (data fetching)
• Plotly (visualization)
• Google Gemini 2.0 Flash + Phidata (AI analysis)
• PyYAML + python-dotenv (configuration)