Focus on your code, not infrastructure, with Cloud Functions
summary
Prototype SMS joke service uses Google Cloud Functions with Twilio to serve replies, starting as a 15-line index.js with three hard-coded jokes and evolving to read a Google Sheet, then export to Cloud SQL and publish analytics to BigQuery via Pub/Sub. Latency drops from ~500ms to ~70–80ms, content editing moves to a spreadsheet, the system scales to thousands of jokes, and cost stays tiny (100k and 1M calls free, ~ $23 at ~10M).
AI-written summary · about
chapters (11)
- 0:00Introduction
- 2:15Story setup: Ingest startup
- 5:05Cloud Functions simplify infra
- 8:21MVP with hardcoded jokes
- 10:30Jokes in Google Spreadsheet
- 16:55Scaling with Cloud SQL
- 20:24Adding analytics
- 23:13Decoupling with Pub/Sub
- 27:04Serverless cost efficiency
- 29:13Cloud Functions use cases
- 30:17Q&A
related talks
54mFocus on Your Code, Not Infrastructure, with Google Cloud Functions
49mGet silly with serverless
36mCode, Build, Run, and Observe with Cloud
35mGoogle Cloud Functions and Firebase
36mWhat you can build with Cloud Functions
50mGo Serverless with Google Cloud
35mConvergence of Serverless APIs & Compute
37mPerformance Tales of Serverless
