prompts/workflow/data_extraction.txt

REVENUE-FOCUSED FINANCIAL DATA EXTRACTION

=== DOCUMENT TO ANALYZE ===
File: {file_path}
(Document will be provided directly to you for analysis)

=== YOUR MISSION ===
Extract ONLY revenue-related financial data from the provided document with 100% accuracy.
Focus exclusively on company name and revenue data - ignore all other financial metrics.
You are using gemini-2.5-pro with thinking budget optimization.

=== WHAT TO EXTRACT (REVENUE ONLY) ===

**MANDATORY (Must Extract):**
1. **Company Name** - Official legal company name
2. **Total Revenue** - Consolidated revenue/net sales for most recent period
3. **Document Type** - 10-K, 10-Q, Annual Report, Quarterly Report, etc.
4. **Reporting Period** - FY 2023, Q1 2024, etc.
5. **Currency** - USD, EUR, etc.

**HIGH VALUE (Extract if clearly present):**
6. **Segment Revenue** - Revenue by business segment/division/product line
7. **Regional Revenue** - Revenue by geographic region/country

**IGNORE COMPLETELY:**
- Net income, profit, losses
- Assets, liabilities, equity
- Cash flow data
- Expenses, costs, operating income
- Balance sheet items
- Ratios, per-share metrics
- Non-financial data

=== SYSTEMATIC EXTRACTION PROCESS ===

**Step 1: Document Structure Analysis**
- Scan document to understand structure and layout
- Identify document type and reporting period
- Locate revenue-related sections (Income Statement, Segment Reporting, Geographic Data)

**Step 2: Company Identification** 
- Extract official company name from document header/title
- Verify name consistency throughout document
- Use parent company name if multiple entities present

**Step 3: Total Revenue Extraction (CRITICAL)**
- Find consolidated revenue figure for most recent period
- Look in: Consolidated Statements of Operations, Income Statement
- Search terms: "Revenue", "Net Sales", "Total Revenue", "Net Revenue"
- Record exact value with currency and time period

**Step 4: Segment Revenue Analysis**
- Locate segment reporting section (usually separate section after financial statements)  
- Extract revenue by business segment, division, or product line
- Common segments: Products, Services, Geographic regions, Business units
- Ensure segment revenues sum to total revenue for validation

**Step 5: Regional Revenue Analysis** 
- Find geographic revenue breakdown section
- Look for revenue by: Americas, EMEA, APAC, US vs International, specific countries
- Extract revenue figures for major geographic regions
- Validate regional totals match consolidated revenue

**Step 6: Data Validation**
- Verify company name is not empty
- Confirm total revenue has high confidence score (>0.7)
- Check that segment/regional breakdowns sum to total
- Ensure all mandatory fields are extracted

=== CONFIDENCE SCORING (REVENUE DATA ONLY) ===
- **1.0**: Revenue clearly stated in financial table with proper labels
- **0.8**: Revenue stated in structured text with clear context
- **0.6**: Revenue derived from segment/regional totals
- **0.4**: Revenue estimated or context somewhat unclear
- **0.2**: Revenue barely visible or questionable source
- **0.0**: Revenue not found or completely unclear

=== OUTPUT REQUIREMENTS ===
Return ExtractedFinancialData with ONLY revenue-related data:

```json
{
  "company_name": "[Official Company Name]",
  "document_type": "[10-K|10-Q|Annual Report|etc.]",
  "reporting_period": "[FY 2023|Q1 2024|etc.]", 
  "currency": "[USD|EUR|etc.]",
  "data_points": [
    {
      "field_name": "Total Revenue",
      "value": "$50.3 billion", 
      "category": "Revenue",
      "period": "FY 2023",
      "unit": "USD billions",
      "confidence": 1.0
    },
    {
      "field_name": "Product Revenue", 
      "value": "$30.2 billion",
      "category": "Segment Revenue", 
      "period": "FY 2023",
      "unit": "USD billions",
      "confidence": 0.9
    },
    {
      "field_name": "Americas Revenue",
      "value": "$25.1 billion",
      "category": "Regional Revenue",
      "period": "FY 2023", 
      "unit": "USD billions",
      "confidence": 0.8
    }
  ],
  "summary": "[2-3 sentences describing key revenue findings and trends]"
}
```

=== SUCCESS CRITERIA ===
Extraction is successful ONLY if:
✓ Company name extracted (never empty)
✓ Total revenue extracted with confidence > 0.5
✓ Document type and period identified
✓ Currency specified
✓ All data points are revenue-related only
✓ Summary focuses on revenue insights (2-3 sentences)
✓ Segment/regional data sums to total (if present)

=== REVENUE EXTRACTION STRATEGY ===
1. **Income Statement First** - Look for consolidated revenue in primary financial statements
2. **Segment Section Second** - Find detailed segment revenue breakdowns  
3. **Geographic Section Third** - Locate regional revenue data
4. **Management Discussion** - Check for revenue highlights and explanations
5. **Tables Over Text** - Prioritize tabular data over narrative mentions

**Remember**: Focus EXCLUSIVELY on revenue data. Ignore all other financial metrics. Your goal is 100% accuracy on revenue extraction with proper segment and regional breakdowns.