Skip to content

AI Tools Integration

The Investment Analysis Framework is designed to be AI-controlled. AI models like Claude and Gemini execute the entire research process autonomously using specialized tools.

🤖 Core Concept: AI Controls Everything

Key Point: The AI model (not the human) runs the analysis, interprets results, and performs research using the available tools.

  • Human Role: Configure investment criteria in YAML files
  • AI Role: Execute systematic screening, analyze results, perform deep-dive research, and generate recommendations

Available AI Tool Categories

The framework provides three categories of tools that AI models can use:

1. 🔍 Systematic Screening Tools

Location: src/invest/ai_tools/*/screening_tools.py

AI models use these tools to: - Execute systematic analysis pipelines on large stock universes - Filter stocks based on quality, value, growth, and risk criteria - Generate screening reports with pass/fail indicators - Rank stocks by composite scores

Example AI Workflow: 

# AI model executes this automatically
run_screening_analysis(config="sp500_full.yaml", save_csv=True)
# Results: AI identifies 25-50 companies that pass all filters

2. 📊 Research and Analysis Tools

Location: src/invest/ai_tools/*/research_tools.py

AI models use these tools to: - Perform deep-dive analysis on companies that pass screening - Gather additional financial data and metrics - Analyze business models and competitive positioning - Generate qualitative insights and research reports

Example AI Workflow: 

# AI model automatically researches promising candidates
analyze_company_fundamentals(ticker="GOOGL")
research_competitive_landscape(sector="Technology")  
generate_investment_thesis(company="Google")

3. 💼 Portfolio Construction Tools

Location: src/invest/ai_tools/*/portfolio_tools.py

AI models use these tools to: - Build diversified portfolios from screened stocks - Optimize position sizing and sector allocation - Generate portfolio recommendations and reports - Track portfolio performance and risk metrics

Example AI Workflow: 

# AI model constructs optimized portfolios
build_portfolio(filtered_stocks, target_size=20)
optimize_sector_allocation(portfolio, max_sector_weight=0.25)
generate_portfolio_report(portfolio)

AI Integration Platforms

The framework integrates with two major AI platforms:

Claude Desktop Integration

Directory: src/invest/ai_tools/claude/

Available Tools: - screening_tools.py - Systematic analysis execution - research_tools.py - Company research and analysis
- portfolio_tools.py - Portfolio construction and optimization - data_tools.py - Data gathering and processing

AI Usage: Claude models can directly access these tools through the Claude Desktop interface.

Gemini AI Integration

Directory: src/invest/ai_tools/gemini/

Available Tools: - screening_tools.py - Systematic analysis execution - research_tools.py - Company research and analysis - portfolio_tools.py - Portfolio construction and optimization
- data_tools.py - Data gathering and processing

AI Usage: Gemini models can access these tools through the Gemini AI platform.

Complete AI-Controlled Workflow

Here's how an AI model would typically use this framework:

Step 1: AI Runs Systematic Screening

# AI model executes systematic analysis
poetry run python scripts/systematic_analysis.py configs/sp500_top100.yaml --save-csv

# AI automatically:
# - Fetches S&P 500 stock data
# - Applies quality, value, growth, risk filters  
# - Generates screening results with pass/fail indicators
# - Identifies 20-40 companies that meet all criteria

Step 2: AI Analyzes Screening Results

# AI model automatically:
# 1. Reads the generated CSV file
# 2. Identifies stocks that passed all filters (Passes_Filters = Y)
# 3. Ranks them by composite score
# 4. Selects top candidates for deep-dive analysis

Step 3: AI Performs Deep Research

For each promising candidate, the AI model: - Uses research tools to gather additional data - Analyzes business model and competitive position - Evaluates management quality and strategy - Assesses industry trends and growth prospects - Generates investment thesis for each company

Step 4: AI Constructs Portfolio

Finally, the AI model: - Uses portfolio tools to build diversified portfolios - Optimizes sector allocation and position sizing - Considers correlation and risk factors - Generates final investment recommendations

Key Advantages of AI Control

🎯 Objective Analysis

  • AI eliminates human emotional bias and confirmation bias
  • Consistent methodology applied to every stock
  • No cherry-picking of favorable data

⚡ Scale and Speed

  • AI can systematically analyze 500+ stocks in minutes
  • Deep-dive analysis on 20-50 companies in hours
  • Human would need weeks for equivalent analysis

🔄 Reproducible Process

  • AI follows identical methodology every time
  • Results are consistent and comparable across time periods
  • No human inconsistencies or varying approaches

🧠 Comprehensive Research

  • AI can simultaneously consider dozens of factors
  • Perfect recall of all relevant information
  • No human memory limitations or oversight

Getting Started with AI Tools

For AI Models (Claude, Gemini, etc.)

  1. Access the Framework: Clone the repository and install dependencies
  2. Choose Configuration: Select or create YAML configuration file
  3. Execute Screening: Run systematic analysis pipeline
  4. Analyze Results: Interpret screening output and identify candidates
  5. Perform Research: Use research tools for deep-dive analysis
  6. Build Portfolio: Use portfolio tools for final recommendations

For Humans

  1. Set Investment Criteria: Define screening parameters in YAML files
  2. Configure AI Environment: Set up Claude Desktop or Gemini integration
  3. Provide AI Context: Share this documentation with your AI model
  4. Let AI Execute: Allow AI to run the complete research workflow
  5. Review Results: Examine AI-generated analysis and recommendations

Detailed Example Prompts

🎯 Stock Screening Prompts

Basic Value Screening: 

I want to find undervalued quality stocks. Please screen the S&P 500 using conservative value criteria and show me the top 15 opportunities. For each stock, explain why it passed the filters and what makes it attractive.

Growth Stock Analysis: 

I'm looking for high-quality growth stocks that aren't too expensive. Please run the analysis and identify technology companies with strong growth but reasonable valuations.

International Opportunities: 

Warren Buffett has been investing in Japan recently. Please analyze Japanese stocks using the buffett_favorites configuration and explain why these companies represent good value compared to US stocks.

Sector-Specific Analysis: 

I want to invest in healthcare but need help finding the best opportunities. Please screen healthcare stocks and recommend companies with strong competitive moats and reasonable prices.

📊 Deep Analysis Prompts

Individual Stock Analysis: 

I'm considering investing in Apple (AAPL). Please run it through your systematic analysis and give me a detailed breakdown:
- Does it pass your quality, value, growth, and risk filters?
- What are its key strengths and weaknesses?
- How does it compare to other technology stocks?
- What's your investment recommendation?

Comparative Analysis: 

Please compare Google (GOOGL) vs Microsoft (MSFT) using your systematic screening. Which one offers better value right now and why? Consider quality metrics, valuation, growth prospects, and risks.

Market Analysis: 

I'm trying to decide between US and international markets. Please analyze both S&P 500 and Japanese TOPIX stocks and tell me:
- Which market offers better value opportunities?
- What are the key differences in quality and growth?
- Should I diversify internationally or focus on US stocks?

💼 Portfolio Construction Prompts

Build Diversified Portfolio: 

Please build me a diversified portfolio of 20 stocks for long-term investing. I want:
- High-quality companies with strong competitive moats
- Reasonable valuations (not overpaying for growth)
- Good sector diversification
- Mix of US and international stocks
- Focus on companies with sustainable competitive advantages

For each recommendation, explain why you selected it.

Value-Focused Portfolio: 

I follow a value investing approach similar to Warren Buffett. Please create a portfolio of 12-15 stocks that:
- Trade below intrinsic value with margin of safety
- Have predictable business models
- Generate strong free cash flows
- Have competitive advantages (moats)
- Include some of Buffett's recent international picks

Conservative Income Portfolio: 

I need a conservative portfolio focused on dividend income and capital preservation. Please find:
- Dividend-paying stocks with sustainable payouts
- Companies with strong balance sheets and low debt
- Businesses in defensive sectors
- International diversification for currency hedging
- Focus on quality over growth

🔍 Research and Analysis Prompts

Industry Deep Dive: 

I'm interested in the semiconductor industry. Please analyze the key players using your screening tools and tell me:
- Which companies have the strongest competitive positions?
- Who offers the best value at current prices?
- What are the key risks and opportunities in the sector?
- Should I invest in chip manufacturers or equipment makers?

Economic Analysis: 

With current market conditions, please analyze what your screening reveals about:
- Which sectors are showing the best opportunities?
- Are growth or value stocks more attractive right now?
- How do US opportunities compare to international markets?
- What does the data suggest about overall market valuation?

Risk Assessment: 

Please run the risk screening and identify:
- Stocks with high debt levels that might be concerning
- Companies with declining business fundamentals
- Overvalued stocks that investors should avoid
- Sectors or regions with elevated risk levels

📈 Follow-up and Monitoring Prompts

Portfolio Review: 

I own [list of stocks]. Please run each through your current analysis and tell me:
- Which ones still meet your investment criteria?
- Have any deteriorated significantly since I bought them?
- Are there better alternatives I should consider?
- Should I rebalance or make any changes?

Market Updates: 

Please re-run your S&P 500 analysis and compare it to your previous screening. What's changed?
- Are there new opportunities that have emerged?
- Have any of your top picks become less attractive?
- How has overall market quality and valuation shifted?

Specific Stock Updates: 

I'm holding Tesla (TSLA) which previously failed your filters. Has anything improved? Please re-analyze and tell me if the investment thesis has changed or if I should still avoid it.

Example AI Conversation Flow

Human: "Please analyze the S&P 500 and find the best value stocks using conservative criteria."

AI Response: 1. "I'll run systematic screening on the S&P 500 using conservative value criteria..." 2. AI executes screening tools 3. "Found 23 companies that passed all filters. Now performing deep-dive analysis on top candidates..." 4. AI uses research tools on promising stocks 5. "Based on my analysis, here are the top 10 value opportunities with detailed investment theses..." 6. AI uses portfolio tools to create diversified recommendations

Human: "Tell me more about the top 3 picks and why they're attractive."

AI Response: AI provides detailed analysis of each company, including financial metrics, business model, competitive advantages, and valuation rationale

Human: "How do these compare to international opportunities?"

AI Response: 1. "Let me analyze Japanese and European markets using the international configurations..." 2. AI runs international screening 3. "Interesting - I found several Japanese companies trading at significant discounts to US equivalents..." 4. AI provides comparative analysis and diversification recommendations

Advanced AI Capabilities

Multi-Modal Analysis

  • AI can process both quantitative data (financial metrics) and qualitative information (news, reports)
  • Combines systematic screening with conversational analysis
  • Generates comprehensive investment research

Dynamic Configuration

  • AI can adjust screening criteria based on market conditions
  • Optimize parameters for different investment styles (value, growth, quality)
  • Create custom configurations for specific investment themes

Continuous Monitoring

  • AI can re-run analysis periodically to track changes
  • Monitor portfolio performance and suggest adjustments
  • Alert to significant changes in screened companies

Next Steps

Remember: This framework is designed to empower AI models to perform comprehensive investment research autonomously. The AI controls the entire process from screening to final recommendations.