MLA-C01 Question 273
Single answerYou are tasked with deploying a machine learning model on AWS to serve predictions for a web application that requires low-latency responses. The application traffic is expected to be unpredictable, with occasional spikes. Which type of endpoint should you choose to meet these requirements?
- A
Serverless endpoint
- B
Real-time endpoint
- C
Asynchronous endpoint
- D
Batch inference
Show answer and explanation
Correct answer: A
Explanation
To meet the requirements of low-latency responses and unpredictable traffic patterns, a serverless endpoint is the best choice. It scales automatically to handle spikes in traffic and removes the need for infrastructure management, making it ideal for such scenarios.
- A. Correct.
Serverless endpoints are well-suited for unpredictable traffic patterns because they automatically scale based on demand and eliminate the need to manage infrastructure. They also support low-latency use cases.
- B. Incorrect.
Real-time endpoints can provide low-latency responses, but they require manual provisioning of the endpoint's infrastructure. This makes them less suitable for unpredictable traffic spikes.
- C. Incorrect.
Asynchronous endpoints are designed for use cases where predictions can be delayed and processed in batches, which is not suitable for low-latency requirements.
- D. Incorrect.
Batch inference is used for processing large datasets in bulk and is not appropriate for serving low-latency predictions in real-time.