Reducing Context Switching & Unplanned Work Through Platform Engineering
summary
Yotpo moves from ~90 weekly DevOps interrupts to ~30–40 by building a layered self‑service platform: Backstage front end, Slack integration that opens Jira tickets, AI triage experiments with enjo.ai, and an internal IDP plus a DevSlab club that gives developers training, permissions and 20% dedicated time. Early pure-AI replies fail because people want humans, but iterative, data‑driven experiments plus cataloging recurring issues cut DevOps requests ~50%, restore developer focus and speed up R&D.
AI-written summary · about
chapters (15)
- 0:00Introduction to context switching
- 1:50Daily chaos of interruptions
- 3:17Measuring and categorizing requests
- 5:03Failed automation attempts
- 6:29Trivial requests and frustration
- 7:33Prioritizing and reducing load
- 9:00Introducing Internal Developer Platform
- 11:11Platform engineering benefits
- 12:17Developer challenges in adoption
- 14:24Managerial benefits of autonomy
- 15:53Cost reduction and agility
- 16:54Dedicated time for platform work
- 17:36Data-driven improvement process
- 18:21Culture of ownership and responsibility
- 19:02Hiring strong engineers
related talks
21mImproving your Developer Experience Through Platforms and Generative AI
29mImproving Developer Experience
11mBuilding Scalable Developer Platforms
54mDeveloper experience platforms
46mForget Developer Platforms, Think Developer Productivity!
18mLeveraging AI to Improve the Developer Experience
10mCombine Support and DevOps work in an Agile team
29mSuccessful Cloud Native Platform Engineering – Hype vs Reality