AI-Driven Developer Productivity at Scale
summary
Practical playbook for immediate use: hook PRs into GitHub Actions to run model-based code reviews against OpenAI or Claude, use Cloud Code or Codeex as agentic CLIs to have an LLM open batched PRs and run compile+tests, and orchestrate agents with an agent plan-execute-verify pattern to reduce hallucinations. Sonnet 3.7/4 shows better self‑timing and reasoning. Real result: a migration pipeline knocks 18 months saved off work and teams target roughly 30% efficiency gains.
AI-written summary · about
chapters (17)
- 0:00AI-driven developer productivity intro
- 3:34Getting started with AI
- 5:21AI in software lifecycle
- 7:49Prompt engineering for PR reviews
- 9:15AI coding assistants
- 12:26Teaching juniors with AI
- 16:58AI for large migrations
- 18:44Supporting juniors in AI era
- 21:13Evaluating AI models
- 24:22Testing AI migrations
- 27:09Human oversight in AI
- 27:30From solo builder to orchestrator
- 29:59Choosing agentic frameworks
- 33:14Measuring AI impact
- 34:37Adopting AI at scale
- 36:22Cost and ROI of AI tools
- 37:02Closing remarks and Q&A
related talks
25mFrom AI Chaos to Production Ready: Using AI Agents to Improve Your Code Quality
32mFrom Toil To Triumph: Harnessing Agentic AI To Streamline Infrastructure as Code
32mBeyond the Hype: A Pragmatic Look at How Developers Actually Use AI Tools
01h 07mNew Methods of Working with AI.
36mAgentic CI/CD: from Pull Request to Production without the Paper Cuts
36mUnleashing Generative AI: Improving Developer Productivity in SRE
41mGoing beyond copilot with AI Agents
26mYour pipeline is the bottleneck