AI / LangGraph LangChain Interview questions
How do you create custom agents?
The easiest way to create a custom agent is with the factory functions create_react_agent() or create_openai_tools_agent(), which combine a custom prompt, an LLM, and a list of tools. Most customisation needs are met by adjusting the prompt and tool list.
from langchain import hub from langchain.agents import create_react_agent, AgentExecutor from langchain_openai import ChatOpenAI from langchain_community.tools.tavily_search import TavilySearchResults # Pull base ReAct prompt from Hub or define your own prompt = hub.pull("hwchase17/react") tools = [TavilySearchResults(max_results=3)] llm = ChatOpenAI(model="gpt-4o") agent = create_react_agent(llm=llm, tools=tools, prompt=prompt) executor = AgentExecutor(agent=agent, tools=tools, verbose=True) executor.invoke({"input": "What is the population of France?"})
For full control, subclass BaseSingleActionAgent (returns one action per step) or BaseMultiActionAgent (returns multiple actions per step). You must implement plan() and aplan() which receive the current intermediate steps and return either an AgentAction (tool to call) or AgentFinish (final answer).
For production multi-step agents with complex state and human-in-the-loop needs, LangGraph's graph-based approach is more appropriate than subclassing agent base classes.
More Related questions...
