Question 1

Which statement best describes the chatbot vs agent roles?

❌ Roles are interchangeable
❌ Rollout agent is generative; chatbot is agentic
✅ Chatbot = generative AI. Rollout system = agentic AI (multi-step autonomous tasks).
❌ Both are generative

Explanation:
Generative AI produces content (scripts).
Agentic AI performs multi-step tasks with autonomy (scheduling, tracking, adapting).

---

Question 2

Core benefit of LangGraph over basic control structures:

✅ Modular nodes + explicit state + conditional transitions for complex agent workflows
❌ Executes all steps in parallel
❌ Simplifies prompting
❌ Eliminates LLMs

Explanation:
LangGraph is built for agent workflows requiring state, branching, and multi-step reasoning.

---

Question 3

Why is LangGraph better than traditional constructs?

✅ State persistence, dynamic branching, human-in-loop support
❌ Built-in ML
❌ Uses fewer resources
❌ Better loop visualization

Explanation:
LangGraph can remember conversation history and change behavior dynamically.

---

Question 4

Key difference between LangChain and LangGraph:

✅ LangChain = linear chains; LangGraph = stateful nonlinear graph workflows
❌ LangChain is not for LLMs
❌ LangGraph replaces components with visuals
❌ LangChain = multi-agent, LangGraph = single-agent

Explanation:
LangGraph is designed for complex agent state transitions; LangChain is mainly pipeline-based.

---

Question 5

Essential step for terminating workflow at n ≥ 13:

❌ ChainState return object
❌ print_out node
❌ Set entry point to END
✅ Use add_conditional_edges with a stop_condition function

Explanation:
Conditional edges determine whether the graph loops or exits.

---

Question 6

Primary purpose of TypedDict in LangGraph state:

❌ Update state dynamically
✅ Explicitly define & type keys used in the state dictionary
❌ Visualize transitions
❌ Auto-convert functions

Explanation:
TypedDict enforces structure, ensuring all nodes use consistent state keys and types.

---

Question 7

Method to specify the initial node:

❌ compile()
❌ invoke()
✅ set_entry_point()
❌ add_node()

Explanation:
set_entry_point() tells LangGraph which node begins execution.

---

🧾 Summary Table