Definition
Sub-agents
Specialized child agents spawned to work in parallel on pieces of a larger task. Mike Vitez reports using more than forty in parallel for one research project.
From the episode “Vibe Coding Your MVP Is a Time Bomb: AI Workflows for Founders Who Want to Ship Twice”Hear it on the podcast1 episode
Learn the language of AI-native founders
Every week: the new terms, tools, and workflows founders are using, straight from the podcast.
Related terms in Agents & Automation
Retro Agent RetirementWhen a redundant agent's job gets reabsorbed by another role and its memory rolls into the organizational record. The agentic equivalent of a restructuring.Routine (automated watcher)An automated watcher, referenced in the episode as R3 and R5, that triggers the next action when a status changes.Self-Healing InfrastructureA system where AI agents detect when a service has failed, identify the root cause, fix it, and deploy the fix — all without human intervention. Used by James Everingham as a 'science fiction challenge' at Meta to drive organic agent adoption.Single-Player AgentAn agent used by one person in a private or local workflow.Swarm of AgentsThe predicted future state of enterprise AI: not one monolithic super-agent, but hundreds or thousands of specialized agents handling specific tasks in parallel. Specialized agents are more debuggable and reliable than any single black-box agent designed to do everything.Vibe codingBuilding software by describing it in natural language to an AI coding tool.Voice AI AgentAn AI system that conducts natural, full-duplex voice conversations with humans in real time - receiving spoken input, processing it through a language model, and responding with synthesized speech in under 500 milliseconds. Distinguished from older interactive voice response (IVR) systems, which present pre-recorded menus and require button presses or simple spoken commands, voice AI agents support free-form conversation, contextual follow-up, and complex multi-turn dialogue. Loman.ai's voice AI agent for restaurants illustrates the commercial use case: it answers the phone as if it were a human employee, knows the full menu and real-time stock status via POS integration, takes orders that print directly in the kitchen, and builds a diner profile from every call. About 40% of callers cannot tell they are speaking to AI - a benchmark that prior-generation voice systems never approached.Wireframe (Legal Context)David's term for using an AI-drafted document as a starting point for attorney review, not as a final legal product. Brings 80% of the drafting work to a lawyer meeting, dramatically reducing billable hours.