Balancing fast-paced innovation and technical debt
summary
Slack operates at scale — 150 million peak messages per minute, 130,000 paid customers, 9 PB of storage and over 2 trillion metrics — and the engineering priority order is concrete: fix customer issues fast, then make sure they don't happen again, then build product. Cycle time is the lever: measure story start to full rollout, break it into write/compile/test/merge/deploy, shave hours off slow code reviews and builds, and set SLAs on remediation items so interrupts drop and velocity rises.
AI-written summary · about
chapters (8)
- 0:00Introduction and context
- 2:35What engineers want
- 4:24The technical debt challenge
- 5:45Prioritizing engineering work
- 8:28Managing incidents effectively
- 12:14Accelerating innovation cycle
- 15:55Conclusion
- 17:07Q&A session
related talks
42mMoving Fast and Securing Things
19mScaling Slack
28mHow we do developer productivity at Slack
45mAccelerating Software Delivery: Why Your Team Is Slowing Down
28mEvolution of Incident Management at Slack
04mHighlights - Future of Work
31mCut me some slack
23mManaging diverse engineers for high-impact deliveries