AI-augmented SDLC: Rethinking how we deliver software
summary
Generative AI capability doubles roughly every seven months but organisations report only a 5% productivity impact because tools amplify both strengths and dysfunctions. Developers spend about 40% of time coding and 60% on context, setup, meetings and debugging so gains at the editor are local. Practical patterns shown: a user story generator using three inputs and LLM eval, test case and data generation that hits ~70–80% of routine work, observability that auto‑diagnoses alerts and proposes MRs for closed‑loop remediation. Measure DORA and business outcomes.
AI-written summary · about
chapters (14)
- 0:00The AI productivity paradox
- 1:23Foundations for AI success
- 3:09Common impediments to flow
- 6:33Slow feedback loops cost
- 8:02Investing in foundations
- 10:27Applying AI across the SDLC
- 12:22AI for user stories
- 14:48Refining user stories
- 16:33AI in coding
- 17:09AI for test generation
- 18:25AI-powered operations
- 19:45Legacy system modernization
- 21:30Measuring AI impact
- 24:10Conclusion and takeaways
related talks
38mFrom code to culture: How AI is reshaping the engineering organization
18mLeveraging AI to Improve the Developer Experience
41mAI-Augmented Testing: How Generative AI, Prompt Eng Turn Testers Into Superheroes
30mIncorporating AI into your SDLC
18mHumans + AI: The Future of Software Development Teams
33mAI assisted software delivery
38mGenerative DevOps
27mAI-First Software Maintenance & AI-First Software Delivery