SaaS release engineering: balancing speed, safety & scale
summary
Platform shows concrete outcomes: merging 100+ PRs and running 500+ builds every two weeks, resolving 500+ vulnerabilities, and shipping 35 major plus 105 patch releases with ~150,000 LOC changed; on-demand micro‑environments (Casmos) cost about $0.24/hr and scale to ~70 concurrent test clusters; CI enforces image scans and policy gates, auto-creates Jira for fixes, and tracks exact release descriptor JSONs for tested configurations; a self-service platform called Sentinel at Acceldata enables one‑click promotions and audited traceability.
AI-written summary · about
chapters (13)
- 0:00SaaS release engineering intro
- 1:31Excel Data’s service infra
- 3:22Release cadence & complexity
- 5:11Enterprise SaaS release challenges
- 6:55Shift-left testing goals
- 8:40Sentinel platform blueprint
- 11:29Micro environments with Cosmos
- 15:21Cost & scalability of clusters
- 16:44Tracking tests with descriptors
- 21:58Vulnerability management in CI/CD
- 26:09Metrics of new system
- 27:11Build vs buy tooling
- 30:40Q&A
related talks
29mAutomating Continuous Compliance for Scalable SaaS Delivery
31mRelease Engineering as a Service - The Art of Integrating with K8s
46mHow to Adapt to Security Changes in Cloud SaaS Transformation
21mPlatform Security
19mApplication Security at High Velocity
53mTransforming testing methodologies: From manual to robust strategies
23mAccelerating SaaS Service Delivery:From 18 Weeks to 3 Hours
26mDevOps and Lean in Legacy Environments