Applications of Foundation Models II
Equip yourself with the necessary understanding to confidently tackle exam questions by reviewing assessment questions and explanations against all options.
We'll cover the following...
Question 34
A legal firm manages a large repository of legal documents, including case law, contracts, statutes, and client files. They want to design a chatbot that allows them to query the repository quickly and retrieve relevant documents based on semantic queries.
As an AI practitioner, which approach would you recommend for training the model with the least operational overhead?
A. Create a knowledge base using Bedrock and Amazon Neptune as a graph-based vector store to store document embeddings and metadata. Use the Bedrock foundational model to process the user query and perform a semantic search to retrieve related documents.
B. Use Amazon SageMaker to train a custom semantic search model on the legal documents and store the embeddings in Amazon OpenSearch Service. Integrate the model with LangChain for querying the document repository.
C. Store legal documents in Amazon DynamoDB and implement a keyword-based search for query handling using LangChain integration. ...