Start As You Mean To Go On
summary
A practical prescription: set alignment first so many teams can move fast within shared rules, then grant autonomy. Uses concrete tech choices — pick Postgres and Kafka for downstream efficiency, prefer JVM-compatible language families over jumping to Go, and design for 10x expected load. Shows real failures: a Citibank screen error that sends roughly $500M out the door, and a telecom that scales from 5,000 to 10,000 people. Build communities of practice, publish narrow policies, and stop over- or under-specifying architecture.
AI-written summary · about
chapters (12)
- 0:00🚀 Build Stuff Conference – Get Your Tickets!
- 2:00Introduction to "Start as You Mean to Go On"
- 4:00Why Architecture Matters from the Start
- 7:00Common Pitfalls in Large Organizations
- 10:00Case Study: Citibank's $500M Error
- 14:00The Dangers of Too Little Guidance in Architecture
- 18:00The Consequences of Too Much Guidance
- 22:00Autonomy through Alignment: A New Approach
- 26:00How to Provide Just Enough Guidance
- 30:00The Role of Architects, Engineering Managers, and Principal Devs
- 34:00Designing for the Future: Preparing for Change
- 38:00Key Takeaways & Conclusion
related talks
44mGetting out of the Starting Blocks
19mEngineering for the Long Game
01h 17mWhen Less Is More: Designing & Decoding Complex Distributed Systems
42mDesigning for failure
14mCreating tools to cross machines and networks
55mGoing Resilient
47mDistributed - of Systems & Teams
24mArchitecture of a Fintech Startup