AI / LangGraph LangChain Interview questions
How do you do batch processing with LCEL?
The .batch() method on any LCEL chain processes a list of inputs and returns a list of outputs. Under the hood, LangChain runs the inputs concurrently using a thread pool (synchronous) or asyncio tasks (async), subject to an optional concurrency limit.
from langchain_openai import ChatOpenAI from langchain_core.prompts import ChatPromptTemplate from langchain_core.output_parsers import StrOutputParser chain = ChatPromptTemplate.from_template("Summarise: {text}") | ChatOpenAI() | StrOutputParser() texts = [ {"text": "Article 1 text..."}, {"text": "Article 2 text..."}, {"text": "Article 3 text..."}, ] # Runs concurrently, returns list in input order summaries = chain.batch(texts) # Limit concurrency to avoid rate limits summaries = chain.batch(texts, config={"max_concurrency": 5})
The async equivalent is .abatch(), which is preferred in async applications:
summaries = await chain.abatch(texts, config={"max_concurrency": 5})
Batch processing is ideal for offline data pipelines: indexing document collections, running evaluations against a test set, or bulk extracting structured data from unstructured text. Results are always returned in the same order as the input list, even if individual tasks complete out of order.
How do you limit the number of concurrent requests when using .batch()?
Are results from .batch() guaranteed to be in the same order as inputs?
More Related questions...
