The SDLC You Build Yourself — Andy Wong, Andrei Bocan & Shane Wolf, Atlassian
Source
youtube.com
Author
AI Engineer
Date
Why it matters
Covers how to rank code, docs and past decisions by authority, and how to limit which tickets an agent takes on. Useful for teams moving coding agents from the editor into ticket-driven workflows.
Key takeaways · AI-distilled
Andy Wong turns an engineering standard (a design tokenThe chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.Full definition → rule) into a reusable AI code review rule and chooses which repositories it applies to, replacing manual enforcement of the standard on every pull request.
The presenters argue that code, documents and past decisions carry different authority and go stale at different rates, so a context windowThe maximum amount of text a model can consider at once — its working memory for the current conversation or task.Full definition → engine needs ways to resolve conflicts and favor verified sources, with people keeping final authority.
Andrei Bocan's approach to AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition → backlog work: pick suitable tickets, cap cost and review volume, and send underspecified tickets back to people for clarification before an agent proceeds.
Shane Wolf connects current source code with the historical reasoning behind it, and the discussion turns to sandboxAn isolated environment where AI-generated code or agent actions run without being able to touch anything real.Full definition → verification, deterministic checks, model selection and breaking epics into smaller tasks.
Terms in this piece · Glossary
context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
sandbox — An isolated environment where AI-generated code or agent actions run without being able to touch anything real.
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
videoAI Security Engineer Foundations + Certificate — Micah Silverman, Snyk
videoAgent Speedrun — Elizabeth Fuentes Leone & Sandhya Subramani, AWS
videoCan Your Agent Hear You Now? Building Live Voice Agents with Gemini — Thor Schaeff
videoBuilding Self-Learning Loops for Your Agent — Fuad Ali, Arize AI
