SEC Filings Intelligence icon

SEC Filings Intelligence

Extract structured financials, risk factors, and executive compensation from SEC 10-K, 10-Q, 8-K filings. The SEC decoder AI agents trust.

SEC Filings Intelligence

AI-generated

Summary

Extracts structured financial, risk, and executive compensation data from SEC filings for specified companies.

Inputs

  • taskMode — Choose the extraction focus
  • tickers — Company stock tickers (comma-separated). NYSE, NASDAQ, etc.
  • dateRange — How far back to search for filings
  • maxFilings — Maximum number of filings to process per company
  • outputFormat — Structure of the output data
  • ciks — SEC CIK numbers (comma-separated). Use when ticker lookup fails.
  • filingTypes — Specific filing types to extract. Empty uses task mode defaults.
  • extractFinancials — Parse income statement, balance sheet, and cash flow
  • extractRiskFactors — Parse Item 1A risk factors with categorization
  • extractMDA — Parse Management Discussion & Analysis
  • extractCompensation — Parse executive compensation tables
  • comparePriorYear — Include year-over-year comparison metrics
  • ragChunkSize — Target token count for RAG chunks (300-2000)
  • baselineRunId — Previous run ID for change detection

Output shape

array of items

Each item contains the transformed filing data; if no filings match a message item is returned; errors may be returned as items when continueOnFail is true.

Examples

Example 1: Extract annual reports for Apple and Microsoft with standard output format

taskMode=annual_report; tickers=AAPL, MSFT; dateRange=1y; maxFilings=5; outputFormat=standard; additionalOptions={extractFinancials=true,extractRiskFactors=true,extractMDA=true,extractCompensation=true,comparePriorYear=true,ragChunkSize=1000}

Example 2: Generate compact RAG chunks for recent quarterly filings of Google using CIK numbers

taskMode=quarterly_report; tickers=GOOGL; dateRange=30d; outputFormat=compact; additionalOptions={ciks=0000789019,extractFinancials=true,extractRiskFactors=true,extractMDA=true,extractCompensation=true,comparePriorYear=true,ragChunkSize=1000}

Links

Discussion