AI Insights & Azure Cognitive Services Integration

Modern business data is not restricted to clean numbers. Enterprises capture millions of lines of unstructured data:

• Customer support ticket transcripts
• Online product reviews and social media feedback
• Survey responses and employee comments

Using AI Insights & Azure Cognitive Services inside Power BI, developers can extract sentiment scores, detect languages, and identify key phrases directly during data transformation!

1. The Pre-Trained AI Capabilities in Power BI

Power BI includes pre-built connectors to Azure Cognitive Services requiring zero machine learning coding:

2. Running AI Insights in Power Query

1. 1
   In Power Query Editor, select your query containing text comments (e.g., CustomerFeedback).
2. 2
   On the Add Column ribbon tab, click AI Insights (or Text Analytics).
3. 3
   Choose the target transformation:

• Score sentiment: Evaluates customer satisfaction.
• Extract key phrases: Extracts prominent subject words.
• Detect language: Identifies the language code (e.g., en, es, hi).

1. 1
   Select the source text column (e.g., ReviewComment).
2. 2
   Click OK. Power BI evaluates the text using Azure AI models and outputs a structured sentiment score column!

3. Visualizing Sentiment & Key Phrases

Once scored:

• Plot average sentiment score over time to monitor customer happiness trends.
• Use Word Clouds or Treemaps to display the most frequent key phrases associated with negative sentiment scores ($< 0.30$), immediately highlighting product defects or shipping bottlenecks!