Refactoring to serverless
summary
Argument: move integration logic out of app code into automation so topology lives in the platform. Demonstrations use AWS CDK (a few lines of TypeScript replace ~450 lines of CloudFormation), replace a Lambda caller with Lambda destination (only fires on async invokes and wraps the payload in an envelope), remove glue Lambdas via Step Functions direct integrations (11,000+ API actions across ~250 services) and publish DB changes with EventBridge Pipes (≈ $0.40–$0.50 per million events); each refactor reduces runtime components, cost, or maintenance but changes invocation semantics.
AI-written summary · about
chapters (15)
- 0:00Introduction and motivation
- 4:47Automation stack decomposition
- 7:23AWS CDK overview
- 10:35Serverless automation focus
- 12:44Refactoring definition and cycle
- 14:13First refactoring example
- 18:39Serverless refactoring motivations
- 21:08Trade-offs and side effects
- 24:00Quality control workflow example
- 26:52Domain event persistence example
- 30:55EventBridge Pipes benefits
- 32:25Integrating refactoring into flow
- 35:20Business-driven refactoring tips
- 37:09Counter-arguments to refactoring
- 38:37Conclusion and resources
related talks
38mRefactoring to serverless
30mEvolving Serverless Architectures
23mSimplifying design complexities using Serverless Architecture on AWS
36mTransforming SaaS with Serverless Architectures
42mBest practices for serverless developers
01h 04mMigrating to Serverless - an experience report
01h 02mFrom Monolith to Serverless
34mDesigning Data Intensive Applications In Serverless Architecture