Which library family is commonly used to build LLM chains?
Why A is the answer
LangChain and similar tools
LangChain and similar libraries join prompts, retrievers and models into chains.
In detail
- LangChain and similar tools can build the app.
- They help compose prompts, retrievers and tools.
- Log the chain with MLflow, so it can be deployed.
Use case
An illustrative scenario, not a real incident.
A team joins a prompt, a retriever and a model into one chain with LangChain, rather than writing the plumbing by hand.
See it in different ways
sequenceDiagram participant C as pyfunc chain participant M as MLflow participant U as Unity Catalog participant S as Model Serving C->>M: Log the model M->>U: Register a version U->>S: Deploy to an endpoint
flowchart LR S0["Prompt"] --> S1["Retriever"] --> S2["Model"] --> S3["Output"]
Key point: LangChain and similar tools compose LLM chains.
Why the other options are wrong
- B. Apache Airflow
- Airflow schedules pipelines. It does not build chains.
- C. Apache Spark MLlib
- MLlib is for classic machine learning at scale.
- D. scikit-learn
- scikit-learn is for classic machine learning.