Keeping engineers sane when everything's on fire: a fintech replatforming tale
summary
Flagstone rebuilds its core cash platform, ships a working replacement and moves £12 billion of client assets during a gruelling 16 months delivery. The new system uses automated pre‑flight checks, a migration dashboard nicknamed Homer’s car to find moveable client slices, a temporary “future flagstone” tribe to eliminate priority conflicts, and pragmatic QA and automation to recover from a brief Marvin verification failure (~98% rejection initially). Result: staged migrations, weekly visible progress and a successful cutover rather than a single big‑bang disaster.
AI-written summary · about
chapters (19)
- 0:00Introduction and speaker background
- 1:51Marvin’s verification failure crisis
- 3:18Origins of the legacy system
- 5:47Why rewrite the platform
- 8:23Commercial debt and mission
- 11:49Team focus and mission badges
- 12:51Avoiding distractions and conflicts
- 14:38Underestimating complexity
- 16:02Daily replanning and agile basics
- 18:28Waterfall with agile slices
- 20:14Homer’s car dashboard explained
- 23:05Daily prioritization and progress tracking
- 25:31Motivation through visible progress
- 26:14Why Marvin failed verification
- 28:21Human behavior disrupting plans
- 29:49Grit and perseverance in engineering
- 31:55AI’s limited impact on the project
- 35:25Three keys to project success
- 37:10People matter most in tough projects
