portfolIQ

Transparency

Screening Methodology

How portfolIQ collects, analyses and classifies financial assets. Every algorithmic decision is documented and audited.

The three methodology pillars

  1. Verified raw data

    Multi-exchange OHLCV aggregated into VWAP consensus. Each data point is timestamped, traceable and referenced to its primary source.

  2. Explainable AI layer

    Claude Haiku/Sonnet generates historised contextual analyses. Reasoning is stored, versioned and auditable.

  3. dbt data engineer pack

    Ready-to-plug Star Schema: conformed dimensions, fact tables and continuous aggregates. Documented columns, surrogate keys, SCD2.