Agentic AI Patterns
summary
Agentic orchestration uses LangChain4j with Quarkus to compose multiple specialized agents into concrete patterns: sequence, loop (max 5 iterations, exit ≥80% score), parallel, routing and supervisor. A reusable planner abstraction lets teams plug custom planners (including a goal‑oriented shortest‑path planner) and resets per invocation. An agentic scope provides shared memory and context injection across agents, a monitoring console generates invocation reports, and demoed flows (bank transfer, medical/legal router, story writer) run end-to-end in released builds (v1.3 → v1.12 monthly).
AI-written summary · about
chapters (16)
- 0:00Introduction
- 1:21Foundations of AI service
- 5:37Key components
- 11:16From single AI to agents
- 13:25Basic orchestration patterns
- 17:39Sequence workflow
- 20:31Loop workflow
- 24:05Parallel workflow
- 25:32Routing workflow
- 29:47Agentic scope in context
- 33:18Supervisor pattern
- 37:16Monitoring agents
- 39:50Pluggable patterns
- 43:03Goal-oriented workflow
- 45:31Additional features
- 46:35Q&A
related talks
51mAgentic AI Patterns
49mAgentic AI Patterns
35mAI Agent Control Logic with Pattern Matching
27mEmerging Patterns Agentic AI
31mAgentic AI architecture patterns
31mBuilding and Orchestrating Real Production-ready Agentic AI Systems
56mDesign Patterns for Multi Agent Systems
46mAgentic Design Patterns Where You Are the Hero