AI-900 Question 242
Select 3A company wants to use Azure's generative AI capabilities to develop a customer support chatbot that generates human-like responses. Which Azure services and features should they consider to properly implement this generative AI workload?
- A
Azure OpenAI Service for access to pre-trained generative AI models like GPT
- B
Azure Cognitive Search to provide relevant data context to the chatbot
- C
Azure Machine Learning to train a custom generative AI model from scratch
- D
Azure Kubernetes Service (AKS) to host the generative AI model
- E
Azure Monitor to track and analyze chatbot performance metrics
Show answer and explanation
Correct answers: A, B, E
Explanation
To implement a generative AI workload for a customer support chatbot, Azure OpenAI Service provides pre-trained generative models, while Azure Cognitive Search ensures the chatbot can retrieve relevant context for accurate responses. Additionally, Azure Monitor helps track performance metrics to ensure optimal functioning. Training a custom model using Azure Machine Learning or deploying the model on AKS is unnecessary in this scenario since pre-trained models are sufficient and do not require containerized deployment.
- A. Correct.
Correct: Azure OpenAI Service provides access to pre-trained generative AI models like GPT, which can generate human-like responses for the chatbot.
- B. Correct.
Correct: Azure Cognitive Search can be used to retrieve relevant data and context, ensuring the chatbot generates accurate and contextual responses.
- C. Incorrect.
Incorrect: While Azure Machine Learning is useful for custom model training, it is not necessary to train a model from scratch when pre-trained generative AI models are available via Azure OpenAI Service.
- D. Incorrect.
Incorrect: Azure Kubernetes Service (AKS) is primarily for deploying and managing containerized applications. It is not a requirement for implementing a generative AI workload using Azure OpenAI Service.
- E. Correct.
Correct: Azure Monitor is essential for tracking and analyzing the chatbot's performance metrics, enabling continuous optimization and reliability monitoring.