LangGraph
You will learn graph-based orchestration for stateful, cyclic agent workflows: typed state, nodes and edges, conditional routing, checkpointing, human-in-the-loop pauses, and multi-agent composition. Interviewers ask because LangGraph answers exactly what a linear chain cannot — loops, retries, approvals, and resuming a run that was interrupted.
part ofAI agent & RAG frameworksoverview, primer and where to startread it →on this pageshowhide
explore
- State Graph Model6 questions
- Conditional Routing5 questions
- Checkpointing & Persistence6 questions
- Human-in-the-Loop5 questions
- Multi-Agent Patterns5 questions
- Streaming & Observability6 questions
questions
page 2 of 2When do multiple LangGraph agents beat one agent with more tools?
basics
~20 sSplit when a single prompt can no longer carry the job: too many tools to select from reliably, genuinely different models or permissions per role, or work that must run in parallel. Otherwise one agent with a good tool set is cheaper, faster and far easier to debug.
When should LangGraph routing live in a conditional edge instead of inside a node?
basics
~20 sMake the branch a conditional edge when someone needs to see, pause, resume, or retry at that decision point: edges are drawn in the graph, land on step boundaries, and split work into separately observable nodes. Keep trivial in-node conditionals inline.
How do you decide what belongs in LangGraph graph state versus outside it?
basics
~20 sPut in state only what a later node must read or what must survive a pause: identifiers, decisions, accumulated results. Keep large payloads, live clients and secrets out — state is snapshotted every superstep, so its size is a per-step cost and everything in it is exposed.
showing 31–33 of 33