Core Concepts
The building blocks behind every verification run — files, datasets, verdicts, and the stage types that combine to produce them.
These pages document the technical concepts shared across the Verification Platform's products and pipelines: how files and datasets are prepared, what a verdict actually means, and how AI, crowd, and web-research stages combine within a single pipeline run.
Files & Enrichment
How files are uploaded, stored, and enriched with metadata before verification.
Datasets
Versioned collections of media files for training and testing verification pipelines.
Verdicts & Confidence Scores
Understanding the five verdict types, confidence scoring, and audit trails.
Synthetic Data Detection
How Crowdee scores the likelihood that content was AI-generated, and where that signal appears in a verification run.
Web Research
How Crowdee performs live web search and crawling as part of source-tracing pipelines.
AI Output Evaluation
Using the crowd to rate the transparency and explainability of AI-generated verification results.
Continuous AI Monitoring
Recurring crowd-sampled quality checks against your pipeline outputs, with drift detection over time.
Verification Contexts
Providing contextual claims to pipelines for richer, more accurate analysis.
Pipeline Stages
How AI and crowd stages are sequenced within a verification run.
How is this guide?