AI & Data Methodology
Prototype policy · counsel and provider review required before production
Purpose and boundary
Apertux organizes educational evidence so users can better understand securities and their existing holdings. It does not select investments, provide personalized allocations, predict outcomes, or execute transactions.
Research structure
Research presents a summary, evidence, quality and valuation context, favorable/base/adverse scenarios, risks, conflicting evidence, uncertainty, missing information, source references, and questions to consider. Recommendation labels, targets, and stop-loss instructions are prohibited.
Data freshness and sources
Every production result must identify its sources and as-of time and distinguish live, delayed, cached, and simulated information. Polygon-backed market prices, charts, and news are labeled live or delayed; portfolio, profile, and scenario context may remain cached or simulated until separate gates are approved.
AI governance
Model, prompt, methodology, source set, validation result, and corrections are versioned. Portfolio credentials and unnecessary personal data are excluded from AI requests. Customer portfolio data may not be used for model training.
Verification
Users should verify important information through SEC EDGAR, official issuer investor-relations pages, Investor.gov, and FINRA BrokerCheck. AI output may be incomplete or incorrect.