fabric-ai-functions

You can apply AI transformations to large-scale tabular data in Fabric notebooks by using pre-built LLM functions that handle sentiment analysis, classification, embeddings, extraction, and summarization with minimal code. It manages concurrency and endpoint interactions automatically.

The best way to run sentiment analysis on a large pandas or PySpark DataFrame is to use AI functions that leverage the synapse.ml.aifunc library. This approach processes thousands to millions of rows efficiently while managing endpoint concurrency.

Yes, you need the synapse.ml.aifunc library to classify text data in Microsoft Fabric. This dependency is required alongside a pandas or PySpark runtime to execute AI functions like classification, extraction, and translation on your tabular data.

Yes, you can use AI functions to summarize lengthy reports and translate text columns at scale. The skill supports summarization and translation transformations, handling everything from thousands to millions of rows with minimal code in Fabric notebooks.

Yes, this approach works with both pandas and PySpark runtimes for data transformation. The skill requires either runtime environment to execute pre-built LLM functions like grammar correction, response generation, and similarity analysis on your datasets.

Limitations of applying AI functions to large datasets in Fabric include dependency on the synapse.ml.aifunc library and the requirement of a pandas or PySpark runtime. Processing scale depends on endpoint interactions and concurrency management within your environment.