Local Differential PrivacyInteractive Demo

Explore how Local Differential Privacy protects individual data while preserving statistical utility. Built with PostgreSQL Anonymizer extension.

SELECT anon.ldp_grrm(value, ε, max_v) FROM table;All demos & mechanism deep dives →

Three Ways to Query

Pre-Anonymized

Query already processed LDP data

On-the-fly

Apply LDP with custom ε or PTTT

Insert & Query

Add your data, see anonymized results

Choose a Scenario

How LDP Works

Epsilon (ε) controls the privacy-utility tradeoff:

  • Lower ε → More privacy, less accuracy
  • Higher ε → Less privacy, more accuracy

Kept Percentage (PTTT) is more intuitive:

  • 65% PTTT → 65% of responses are truthful
  • Automatically converted to equivalent ε