how it works • transparency lab

collect statements

we track every public statement, promise, and prediction from verified sources. timestamped, sourced, and catalogued.

verify outcomes

cross-reference with financial filings, production data, and documented results. no interpretation—just measurable facts.

detect contradictions

algorithmic analysis finds conflicts between statements, promises vs. outcomes, and timeline divergences.

visualize connections

present everything in interactive formats. timelines, relationship graphs, explorable datasets. you draw conclusions.

contradiction detection

critical

directly contradictory

two statements that cannot both be true. example: "we will deliver 500k units" vs. actual delivery of 250k units.

high

timeline failures

promises made with specific timelines that were not met or significantly delayed.

medium

narrative conflicts

statements that tell different stories about the same event or metric across time.

low

minor inconsistencies

small discrepancies that may have reasonable explanations but are worth noting.

data sources

statements: twitter/x, earnings calls, interviews, press releases, regulatory filings

outcomes: sec filings (10-k, 10-q, 8-k), production reports, financial statements

relationships: org charts, investment data, board positions, company ownership

all sources are linked and verifiable. no anonymous sources. no speculation.