AI-103: Text Analysis
Text Analysis is one of the skill areas tested on the Developing AI Apps & Agents on Azure (AI-103) exam. Below are free AI-103 practice questions with worked answers and a concept diagram — each with a plain-language explanation. Practice the first 10 questions of AI-103 free, no signup.
The concept, in one picture
2 free AI-103 Text Analysis questions
A Microsoft Foundry agent uses Azure Speech in Foundry Tools. You fine-tune a baseline speech-to-text model for the en-us locale and publish it. When the agent calls the Speech to text REST API, it returns an error saying the project ID is invalid. You need to point the project property at the right identifier. What should you set the project property to?
- the custom speech endpoint URL
- the project URL
- the project ID
- ✓ the custom speech project ID
When using the Speech to text REST API with a fine-tuned custom model, the project property must reference the Custom Speech project, which is a different identifier from the Microsoft Foundry project. Because the custom speech project ID is not the same as the Foundry project ID, supplying the Foundry ID, a project URL, or the endpoint URL triggers the invalid project ID error. The custom speech endpoint URL is only for calling a deployed custom model endpoint during recognition, not for the project property value.
Your Azure Speech in Foundry Tools resource hosts a custom speech-to-text model that is published to a custom endpoint, and an agent calls that endpoint for live speech recognition. The custom model is about to reach its expiration date. Once the model expires, what happens to real-time recognition requests sent to the endpoint?
- ✓ The requests are automatically served by the latest available base model for the matching locale.
- The requests keep using the expired custom model until an administrator manually deletes it.
- The requests fail with a 4xx error until an administrator publishes a replacement custom model.
- The expired custom model is automatically purged from the resource.
After a custom speech-to-text model expires it can no longer be used for transcription. On a custom endpoint doing real-time recognition, Azure Speech falls back to the newest base model for the same locale, so recognition still works but loses the domain-specific tuning the custom model provided. This is unlike batch transcription, where naming an expired model returns a 4xx error. Expiration neither deletes the model nor keeps the expired custom model in use. The recommended practice is to redeploy the endpoint with a newer custom model before the expiration date.
Practice AI-103 free
The first 10 questions of every exam are free. No signup, no email wall.
Start practicing →Get a free AI-103 study plan by email
A short plan to work through AI-103 by skill area, plus a note when we add new questions. Optional — the practice above stays free. No spam, unsubscribe anytime.