High performing engineering teams and the Holy Grail
summary
CircleCI analyzes over 15 million workflows and claims four actionable CI/CD metrics drive high performance: duration, MTTR, success rate, throughput. Benchmarks shown: ≤10 minute builds (50% at 3.3 min, median ≈11 min due to long tails), MTTR ≲60 minutes (median ~64), default-branch success target ≥90% (observed ~77%), and median throughput ~1.5 runs/day. Advises stronger test coverage, caching dependencies, SSH into failed builds, and creating a Platform Team / IDP to abstract ops, remove blockers, and improve these specific numbers.
AI-written summary · about
chapters (12)
- 0:00Introduction and welcome
- 1:35Why engineering teams matter today
- 3:23Economic impact on software delivery
- 5:55Platform teams and their role
- 7:19Four key engineering metrics
- 8:21Optimizing pipeline duration
- 12:57Improving test coverage
- 14:21Mean time to resolve failures
- 19:14Success rate benchmarks
- 22:45Understanding throughput
- 27:01DevOps vs platform engineering
- 32:25Q&A
related talks
25mContinuous Delivery, high-performing engineering teams, and the Holy Grail
50mFast Doesn’t Mean Fragile: Delivering AI-Powered Software at Scale
27mThe Need for Speed (in your CI/CD setup)
48mDemystifying CI/CD
20mMaximize development efficiency with expert CI/CD strategies
24mCI/CD Best Practices to Deliver Software Faster
24mWhat a global pandemic can tell you about better DevOps practices
46mThe Need for Speed (in Your CI/CD Setup)