Applied Agentic AI

Practical, production-oriented implementations of Agentic AI systems, focusing on autonomous planning, execution, validation, memory and multi-agent collaboration.

---

What is Agentic AI?

Agentic AI systems are capable of:

• Planning tasks autonomously
• Executing actions using tools
• Maintaining internal state and memory
• Validating and refining outputs
• Working toward long-horizon goals

---

Projects

#	Project	Notebook	Description
1	Autonomous Task Planner Agent	agentic_task_planner.ipynb	A goal-driven agent that decomposes high-level objectives into actionable steps and executes them using a plan–execute–verify loop.
2	Multi-Agent RAG System	multi_agent_rag_langgraph.ipynb	A multi-agent retrieval-augmented generation system where retrieval, answer generation, and critique are handled by separate agents to reduce hallucinations.
3	Self-Reflecting Agent with Memory	self_reflecting_agent_langgraph.ipynb	An agent that critiques its own outputs, records failures, and improves responses over time using reflection memory.
4	Tool-Using Agent	tool_using_agent_langgraph.ipynb	An agent that decides when to invoke external tools, executes them, and integrates tool outputs into final responses.
5	Re-Ranking Agent in RAG	reranking_agent_rag_langgraph.ipynb	A RAG system enhanced with a re-ranking agent to refine retrieved documents and improve answer relevance.
6	Long-Horizon Goal Agent	long_horizon_goal_agent_langgraph.ipynb	An agent that decomposes long-term goals into milestones, tracks progress, and adapts plans based on feedback.
7	Persistent Memory Agent	persistent_memory_agent_langgraph.ipynb	An agent with vector-based persistent memory that retrieves past experiences to influence decisions across runs.
8	Multi-Agent Collaboration System	multi_agent_collaboration_langgraph.ipynb	A system of specialized agents that collaborate through planning, execution, and review to produce higher-quality outcomes.
9	Self-Evaluating Agent with Metrics	self_evaluating_agent_langgraph.ipynb	An agent that evaluates its own outputs using explicit quality metrics and refines responses based on measured performance.
10	Semantic Graph Memory Agent	semantic_graph_memory_agent_langgraph.ipynb	An agent that stores knowledge as a semantic graph (nodes and relationships) and answers questions by reasoning over structured memory.
11	Enterprise Agent Orchestrator with Observability	enterprise_agent_orchestrator_langchain.ipynb	A production-grade LLM orchestration pipeline using LangChain’s ChatOpenAI with environment-based configuration, retry logic, model fallback routing, cost tracking, and structured observability.
12	Supervisor-Controlled Routing Architecture	supervisor_controlled_routing_langgraph.ipynb	A LangGraph-based supervisor architecture where a central governance agent dynamically routes tasks to specialized agents based on intent classification and policy rules.
13	Planner–Executor–Verifier Architecture	async_planner_executor_verifier_langgraph.ipynb	An asynchronous LangGraph-based Planner–Executor–Verifier agent architecture.
14	Event-Driven Agent Architecture	event_driven_agent_architecture_langgraph.ipynb	A LangGraph-based event-driven agent system where agents react to emitted events through conditional routing.
15	Blackboard Multi-Agent Architecture	blackboard_agent_architecture_langgraph.ipynb	Independent specialist agents collaborate by reading and writing to shared memory instead of directly invoking one another.
16	Distributed Agent Execution	distributed_agent_execution_simulation_langgraph.ipynb	A LangGraph-based distributed agent execution simulator that parallelizes task processing across worker pools using asynchronous orchestration.

---

Notes

• All projects are implemented using LangGraph for explicit control flow and state management.
• The focus is on reliability, explainability and production-oriented agent design, not just prompt chaining.