MLOps.community Spec Driven Development, Workflows, and the Recent Coding Agent Conference
21 snips
Mar 31, 2026 Jens Bodal, a senior software engineer who builds backend systems and developer tooling, discusses how abstract AI agents shift work from coding to defining intent. He covers agent orchestration, evaluation challenges, security and sandboxing, local self-hosted stacks, spec-driven workflows, and how teams must rethink ownership, reviews, and telemetry in an agent-first world.
AI Snips
Chapters
Transcript
Episode notes
When Specs Replace Reading Code
- Spec-driven development can replace reading code when specifications are detailed enough to prevent agent derailment.
- Jens Bodal warned a single typo in a spec can derail the whole system, so specs must be precise and tested.
Agent Vocabulary Is Still Unsettled
- Agent terminology (agent, subagent, co-agent, ephemeral) is fluid and causes communication friction.
- Jens defines subagents as subprocesses and describes adversarial/ephemeral agents used as judges in orchestration loops.
Evaluate Agents On More Than Accuracy
- Evaluate solutions across multiple axes: accuracy, cost, latency, and engineering effort.
- Jens contrasts vector search and AST/LSP-based code-edit approaches to show trade-offs between resource cost and correctness.