How big companies actually deploy AI agents: the agentic app pattern
Cole Medin · youtube · 2026-10-09
Cole Medin shares lessons from deploying AI agents at enterprises, shaped around Oracle's "agentic apps" pattern — quite different from a single chat agent:
- Multiple agents working toward one goal instead of one chat box
- LLM calls only where judgment is needed; everything else stays deterministic
- Rules turned into actual code with tests, not left in prompts
- Human approval for anything that matters
He demos building a leave-policy workflow in Oracle AI Agent Studio: every LLM call runs inside a workflow, policy nodes handle security instead of RAG, with workflow permissions, triggers, and visual outputs rather than walls of text.
More from coding & agent
- Claude Code Pushes Urgent Security Update Across All Installations — umeshonai · 2026-10-23
- Claude cut a 2.6-hour video export to 20 minutes — but only after being told to rebuild the server — byhewar · 2026-10-23
- 3 flight-booking rules my AI agent checks that I'd never do by hand — Efistoffeles · 2026-10-23
- OpenJet claims Qwen3.8-27B coding runs 1.9x faster than llama.cpp on one RTX 3090 — JeremyCMorgan · 2026-10-23
- Building an autonomous SEO agent in Hermes with three skills: programmatic pages, Keywords Everywhere, DataForSEO — EXM7777 · 2026-10-23
- Building a sports betting AI agent with onchain execution and human approval by default: architecture and lessons — Dry-Comfortable-2678 · 2026-10-23