Update responsible-ai-best-practices-genai-prompt-skill.md

Author: JamesJBarnett

articles/search/responsible-ai-best-practices-genai-prompt-skill.md

@@ -27,7 +27,7 @@ In order to list out the various challenges in incorporating AI content generati
 
 | Persona | Description |
 |---------|-------------|
-| End-user | The person asking questions of the RAG application, expecting a well-cited answer to their question based on results from the source document. In addition to accuracy of the answer, the end-user expects that any citations provided by the application make it clear if it was from verbatim content from a source file or an AI-powered summary from the model. |
+| End user | The person asking questions of the RAG application, expecting a well-cited answer to their question based on results from the source document. In addition to accuracy of the answer, the end-user expects that any citations provided by the application make it clear if it was from verbatim content from a source file or an AI-powered summary from the model. |
 | RAG application developer/search index admin | The person responsible for configuring the search index schema, and setting up the indexer and skillset to ingest language model augmented data into the index. GenAI Prompt custom skill allows developers to configure free-form prompts to several models hosted in AI foundry, thereby offering significant flexibility to light up various scenarios. However, developers need to ensure that the combination of data and skills used in the pipeline doesn't produce harmful or unsafe content. Developers also need to evaluate the content generated by the language models for bias, inaccuracies, and incorrect information. Although this task can be challenging for documents at a large scale, it should be one of the first steps when building a RAG application, along with the index schema definition. |
 | Data authority | The person expected to be the key subject matter expert (SME) for the content from the data source. The SME is expected to be the best judge of language model powered enrichments ingested into the index and the answer generated by the language model in the RAG application. The key role for the data authority to be able to get a representative sample and verify the quality of the enrichments and the answer, which can be challenging if dealing with data at large scale. |
 
@@ -138,4 +138,4 @@ Given the scale of data ingestion, it might not be feasible to have a human in t
 
 * [Introduction to Azure AI Search](search-what-is-azure-search.md)
 * [AI enrichment concepts](cognitive-search-concept-intro.md)
-* [Retrieval Augmented Generation (RAG) in Azure AI Search](retrieval-augmented-generation-overview.md)
+* [Retrieval Augmented Generation (RAG) in Azure AI Search](retrieval-augmented-generation-overview.md)