Valid AI-103 Dumps shared by EduDump.com for Helping Passing AI-103 Exam! EduDump.com now offer the newest AI-103 exam dumps, the EduDump.com AI-103 exam questions have been updated and answers have been corrected get the newest EduDump.com AI-103 dumps with Test Engine here:
You have a Microsoft Foundry project that contains an agent. The knowledge source for the agent is a set of scanned PDF troubleshooting guides stored in Azure Blob Storage. The guide pages contain two-column layouts and tables. You use Azure Content Understanding in Foundry Tools to process the PDFs. You plan to ingest the processed content into an index for Retrieval Augmented Generation (RAG) and store extracted fields for downstream automation. Stakeholders must be able to verify where each extracted field value came from in the original PDF and route low-reliability extractions for manual review. You need to ensure that the Content Understanding document analyzer output includes a per-field confidence score and source grounding locations within the source document. What should you do?
Correct Answer: A
The correct answer is A. Enable estimateFieldSourceAndConfidence . Azure Content Understanding document analyzers support an opt-in confidence and grounding capability for field extraction. Microsoft documentation states that to opt in for confidence and grounding, you set estimateFieldSourceAndConfidence = true in the analyzer configuration, or configure estimateSourceAndConfidence = true for specific fields. This enables each extracted field to include a confidence score and references back to the original document source location. This directly satisfies both stakeholder requirements: source grounding allows users to verify where the extracted value came from in the scanned PDF, and the confidence score supports downstream automation rules, such as sending low-confidence fields to manual review. Microsoft's analyzer improvement guidance describes confidence scoring as a value between 0 and 1 and grounding as references or citations for extracted outputs to the original document content. Generative extraction does not guarantee per-field confidence and source grounding. enableSegment is used for document segmentation, not confidence scoring. Labeled samples can improve extraction quality, but they do not by themselves enable confidence and grounding output. Reference topics: Content Understanding document analyzers, field extraction, confidence scoring, grounding, and RAG ingestion.
