TechPrufer
Get started
Dropbox
Dropbox logoDropbox · Feature trail

Cookie Compliance Auditing

3 posts · May 2021 – Aug 2026

What this trail covers

Dropbox developed an in-house cookie auditor to ensure consistent and compliant cookie behavior across hundreds of web surfaces. The system automates the testing of cookie banners and Global Privacy Control (GPC) signals, simulating user interactions to verify that privacy preferences are respected. It includes a URL detection mechanism to identify all relevant web properties for auditing, enabling continuous validation of privacy commitments as the web presence evolves. This post details the ap

All Dropbox trails
  1. Testing cookie behavior across hundreds of web surfaces with our in-house auditorThis post details the development of a cookie auditor using Playwright for browser automation to test cookie banner compliance across hundreds of Dropbox web surfaces. It describes the process of defining testable rules from privacy requirements, simulating user interactions (US, EU, GPC signals), and implementing a companion URL detector to identify all relevant web properties for auditing. The system allows for continuous validation of privacy commitments as the web presence evolves.
  2. Meet Chrono, our scalable, consistent, metadata caching solutionThis post introduces Chrono, a scalable, consistent metadata caching solution built on top of Dropbox's Panda key-value storage system. Chrono addresses the challenge of serving high read query volumes while maintaining read-after-write consistency, a critical requirement for internal metadata clients. It achieves this by decoupling write attempt tracking (Chrono Server) from data caching (Memcache). Chrono Server provides an `Attempt` API to notify of write attempts and a `LatestAttemptTimestamp` API to retrieve the highest observed attempt timestamp for a key. This information is used by the Memcache layer to detect and avoid serving stale data, ensuring clients receive up-to-date metadata.
  3. Optimizing payments with machine learningIntroduced machine learning, specifically a gradient boosted ranking model, to optimize payment renewal attempts. The model predicts the best times to charge customers during subscription renewals and retries, moving away from static, rule-based logic. This aims to increase payment charge success rates, reduce collection time, and minimize involuntary churn. The system was integrated with the ML Platform's Predict Service for low-latency predictions.