Automatically generate Backstage entity descriptors for easy adoption
summary
Backstage adoption stalls at ~10% because teams avoid creating manual catalog descriptors; generating them from existing sources (GitHub metadata, CI/CD config, codeowners, runtime deployments) and storing output centrally (example: an S3 bucket) produces YAML entities Backstage ingests. A lightweight entity-descriptor builder that merges repo, infra-as-code and platform data pushes adoption to ~90%, reduces manual errors, and recovers the productivity delta companies report (roughly equivalent to three full-time developers and halving delivery slowdowns).
AI-written summary · about
chapters (10)
- 0:00Internal dev portals intro
- 4:25What is Backstage
- 8:20Empty catalog at Acme Payments
- 13:19Why adoption fails
- 15:26Automating entity descriptors
- 17:59Meta info needed
- 20:54Data sources for automation
- 23:42Benefits of automation
- 25:25Acme Payments success
- 27:29Key takeaways
related talks
49mA Front Row Seat to Backstage Adoption
26mAn Easier Path to a Rich Catalog
32mIs Backstage for me?
35mImproving Developer Experience with Open Source 1
25mHow to Adopt Backstage: Lessons from 20 Adopter Interviews
29mHow We Made Backstage Improve Developer Efficiency of 1000+ En...
01h 01mClosing the documentation gap with Backstage.io
33mThe State of Backstage in 2025