The Executive Branch

WhiteHouse.wiki

Methodology

How WhiteHouse.wiki is built.

This page explains how WhiteHouse.wiki collects executive-branch data, where AI is used, and what guardrails keep the site tied to official records.

Public-interest data

Official sources with visible AI boundaries

WhiteHouse.wiki explains what is source data, what is enriched, and where each page came from.

Sources

Official records come first.

WhiteHouse.wiki starts with official White House publications, Federal Register records, nomination data, and administration roster data collected into the shared Congress.wiki database. Each public page keeps the source link visible so readers can inspect the original publication directly.

Structure

The site connects people, offices, documents, and actions.

Instead of storing the White House as separate feeds, the product links the same records across pages. A document can connect to a topic. A topic can connect to a presidential action. An office can connect to the current leader and historical staffing snapshots. The goal is to help readers follow relationships instead of hunting across disconnected releases.

AI use

AI helps explain and classify, not invent facts.

AI is used narrowly for cheap, fast interpretation work such as short summaries, why-it-matters text, and topic classification on ambiguous policy documents. Core facts such as titles, dates, people, offices, and official source URLs come from deterministic collection and parsing, not model generation.

Quality controls

Prompt changes are tested before full reruns.

White House topic classification is evaluated against committed end-to-end samples before prompt or taxonomy changes are rolled across the full dataset. Deterministic rules still handle obvious or ceremonial cases so the AI pass stays focused on the documents where it actually improves quality.