Project ideas for Azure AI 102 preparation😍

I have to prepare for this certification and, to be honest there is nothing more boring than just sitting around reading about the portal and about how things work without actually coding.

So, I got this idea to build a few microservices (or tools) in C#, that would help me learn things in the practical way.

Here I am listing a few of these topics. If you click the title, it will lead to the gitHub repository containing the solution. Also, some of them are in the same tool so don’t be surprised if they redirect you to the same resource.

You can check the readME.md file or the list of api’s in the Program.cs file where the minimal api’s are listed.

Later, we will be creating also the articles that explain everything behind the code. So, without further ado, let’s jump to the plan.

1. AI-Powered Content Moderator API
   • Input: Text (string).
   • Output: Moderation result (offensive/acceptable + category).
   • Uses: Azure Content Safety API.
   • Extras: Store logs in CosmosDB or SQL, versioning of flagged content.
2. Text Sentiment Analysis Microservice
   • Input: List of comments (JSON).
   • Output: Aggregate sentiment scores (positive, neutral, negative).
   • Uses: Azure Text Analytics.
   • Extras: Add filters (date/user) and a summary endpoint.
3. Named Entity Recognition (NER) Extractor
   • Input: Text document.
   • Output: Entities (person, location, organization, key phrases).
   • Uses: Azure Language Service – Entity Recognition.
   • Extras: Optionally store results in MongoDB for querying.
4. Summarization & Translation API
   • Input: Long text (article, document).
   • Output: Summary + translation (to language X).
   • Uses: Azure Document Intelligence / Translator API.
   • Extras: Support multiple languages, return word counts.
5. Image Analysis REST API
   • Input: Image URL or file.
   • Output: Tags, captions, and objects detected.
   • Uses: Azure Computer Vision API.
   • Extras: Optionally, cache results in Azure Blob + SQL.
6. Speech-to-Text Transcription Service
   • Input: Audio file.
   • Output: Transcribed text with timestamps.
   • Uses: Azure Speech Service.
   • Extras: Save transcripts to storage, return text as JSON.
7. Knowledge Base Q&A API
   • Input: User’s question.
   • Output: Best answer from a knowledge base.
   • Uses: Azure AI Search + Language Studio (QnA Maker successor).
   • Extras: Implement feedback loop (store questions that got no good answer).

Why this helps for AI-102

• All these projects map to exam topics: Language service, Vision, Speech, and Responsible AI.
• Practice calling Azure AI endpoints securely from C#.
• Learning how to structure backend APIs for AI consumption (which is a big part of the role).