AI is shifting software development from “write code” to “run systems that write code.” In this episode, Mitch Ashley and Brad Shimmin unpack how AI agents are changing the development workflow, from planning and specifications to execution and maintenance.
They dig into why context is now a core requirement for successful projects, and why teams are hitting friction when existing tooling and APIs can’t support structured, predictable agent interactions. The conversation also covers the practical challenges of memory and state in AI-driven systems, including what happens when an agent can’t retain the right details across a longer project timeline.
Mitch and Brad also discuss what this means for developer roles. The work is shifting toward orchestration, validation, systems thinking, and building guardrails that keep AI productive instead of chaotic. The takeaway is simple: the teams that adapt their process and tooling to agentic workflows will move faster, ship cleaner, and waste less time fighting their stack.