最佳实践

本节包含有关如何有效使用上下文基础的信息。

入门指南

To start using Context Grounding, follow these steps:

1. 确保您使用 Studio Web 或 Studio Desktop 2024.4 或更高版本。
2. In Automation Cloud^TM, in Integration Service, establish a connection to the UiPath GenAI Activities connector.

   This allows you to use the GenAI Activities in Studio.

3. To create and manage indexes, access the Context Grounding tab in Automation Cloud™ > Admin > AI Trust Layer. You can then query indexes from Autopilot for everyone or from the GenAI Activities.

Context Grounding in UiPath GenAI Activities

要使用上下文基础，您需要现成的数据集（例如，一组文档）。 然后，上下文基础可以分三个阶段与您的数据交互：

1. 为上下文基础建立数据源。
   • Context Grounding follows shared folder permissions. Use a folder with appropriate access to manage and query from data.
   • Create a connection to the supported Integration Service data sources or add data to a shared Orchestrator bucket location.
2. 将数据源中的数据提取到上下文基础中。
   • Use the Context Grounding tab in the AI Trust Layer admin to create, sync, and delete indexes.
   • Use the Update Context Grounding Index activity to sync an index you have created. This allows non-admins to help manage the data freshness of an index based on new data.
3. 使用您的数据进行查询和地面提示。
   • Use the Content Generation activity or Autopilot for everyone to query over documents and use information to augment or ground prompts.

管理提取管道

You can manage the ingestion pipelines in two ways:

• Through the Context Grounding tab in AI Trust Layer. See Governing contextual data for GenAI features.
• Through the Update Context Grounding Index activity, part of the UiPath GenAI activities package.

Querying data with Context Grounding

The Content Generation activity features two options for working with documents: File resource and Existing index.

文件资源

The File resource option allows you to use file variables from previous activities (like a document downloaded from Google Drive) and have Context Grounding perform a just-in-time (JIT) style of RAG. This means it ingests the document into an index, searches across it, augments the prompt, and then deletes the index, so the data is not persistent.

备注：

• 保持这些文档较小（不超过 50 页）。 Integration Service 活动设置一个超时窗口，上述所有处理操作均会在该窗口内进行。
• 如果您有包含扫描图像的 PDF，我们建议您在下载文件以从这些图像中提取文本后，在“提取数据”活动中使用 Document Understanding OCR 选项（因为“上下文基础”尚不支持图像）。 将该extracted_data输出传递到提示，并将该文件资源指向您下载的文件。

现有索引

The Existing index option allows you to use an index you have previously created through AI Trust Layer. You query across a persistent index into which you've ingested documents from your data source. You can re-use this index as many times as you like until you delete it.

Use the Log Message activity after Content Generation in the workflow sequence to gather detailed information:

• Input the Top Generated Text or Citations string output variables from Content Generation to see the LLM generation response after the workflow execution.
• Input the Citations string variable to see the semantic search results used to influence the generation output. This works only for PDF and JSON data types.