From Firefighting to Flow
summary
Joined a wholesale fiber provider in Iceland in early 2025 to build a software team and architecture that stop the firefighting loop. Implements limit work in progress and protects midweek improvement time, shifts metrics from tickets closed to lead time, predictability and frequency of surprises, and enforces ownership so three-person ownership replaces heroics. Adds a reverse proxy and local data validation to decouple integrations. A recent release exposes an 18‑year bug where deactivated connections never released equipment and the team fixes the root cause instead of layering another patch.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:21Firefighting chaos
- 4:59System problems behind failures
- 7:08Cost to investment mindset
- 8:56Space for improvements
- 10:43Limit work in progress
- 11:27Redefine success metrics
- 12:10Designing for flow
- 13:32Small frequent changes
- 16:03Eliminate hero culture
- 17:27Psychological safety
- 18:31Mindset and conversations
- 19:57Resilience through prevention
- 20:18Break firefighting cycle
related talks
40mFrom Strategy to Flow: A Journey Using DDD to Achieve Flow
34mFrom Waterfall To River – A Story Of Transitioning To Flow
42mFluid Scaling Technology: When you want to go FAST!
40mControlling Wildfires While Only Getting Singed.
29mIntentional Architecture in Agile Organizations
47mFighting Software Entropy
28mPractical advice for technical decision-making while evolving systems in place
10mI am done -