Technical Metamorphosis: The 3 Stages of DSAR Automation
summary
Uber builds an in-house DSAR pipeline that converts agent selections into queued exports, then executes them as parallelized, recoverable units called cuboids so failures only retry the failed subset. Core pieces are an API layer plus metadata store and a time-triggered batch orchestrator, templated data sources for fast onboarding, and checkpointing to avoid rework. Solution scales by tuning users/time/data per cuboid and cuts engineering hand-holding compared with manual SQL/API fulfillment.
AI-written summary · about
chapters (11)
- 0:00Introduction and welcome
- 1:08What is a DSAR
- 4:14Manual DSAR process challenges
- 6:27Building vs buying solutions
- 8:16Why automate DSARs
- 8:57Automation requirements
- 10:04Automated DSAR tool demo
- 13:40Scaling with cuboids
- 16:39Privacy Center integration
- 17:42Key takeaways and next steps
- 18:50Q&A session
related talks
11mBuilding a Complete Export Ecosystem—From DSAR Automation to Privacy Center
10mData Access Automation at Scale
33mAutomation Best Practices for SRE and Security
59mBuilding resilient event-driven architectures
01h 00mAutomating Security in Your Data Pipeline
36mSecurity Automation In A Devops World
25mBeyond Human Scale: AI and Automation as Force Multipliers in Cyber Threat Intelligence
33mMetabadger: Automating IMDS Protection at Scale in AWS