Continuous Delivery for Python with Feature Flags
summary
Claim: feature flags let teams push incomplete Python code and toggle behavior at runtime for safe rollback and A/B experiments. Demo uses Split.io with a FastAPI demo that adds an address field behind a flag and flips it on in production without restarting the server, proving live toggles work. Also shows control treatment when SDK is unavailable. Notes real costs: added latency on flag fetch and accumulating technical debt from forgotten toggles and nested conditions.
AI-written summary · about
chapters (7)
- 0:00CI/CD and feature flags intro
- 4:12Progressive exposure techniques
- 8:30Feature flag use cases
- 16:59Managing flags with Split.io
- 17:21Best practices for flags
- 22:42Live Python flag demo
- 28:49Q&A
related talks
08mThe Art of Feature Flagging
41mFeature Flags: Separate Deployment from Feature Release
50mThe Story of One Feature Flag
44mFlags of Flexibility - Unlocking Rapid Iteration
01h 09mFeature flags: the toggle, the A/B test and the canary
59mFaster & less risky releases with feature flags
11mFeature flag use cases you haven't heard about yet
50mFeature Toggle Promotion via CI/CD Pipeline