Back to Portfolio
Content Safety Project

Content Safety Analyzer

Paste any text or load content from a URL to analyze it for toxicity, hate speech, threats, and policy violations. Powered by the Hugging Face Inference API and a custom policy enforcement engine.

Content ModerationHugging Face Inference APIToxicity ClassificationPolicy Enforcement EngineNLP / ML PipelineEvidence ExtractionRisk ScoringSeverity ClassificationCheerio HTML ParsingURL Content Analysis

The Hugging Face Inference API lets you call pre-trained ML models via HTTP — no GPU, no server setup, no API key required for free-tier access. This tool uses the unitary/toxic-bert model to classify text across 6 toxicity categories, then maps the scores to policy violations with severity ratings and enforcement actions.

0 / 5000