AI-102 Implement knowledge mining and information extraction solutions — 100 Questions

Question 1

You are building a knowledge mining solution to extract insights from a large set of PDF contracts. The solution must identify parties, dates, and monetary amounts. Which Azure AI service should you use as the primary extraction engine?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Azure AI Document Intelligence. Azure AI Document Intelligence (formerly Form Recognizer) is the correct choice because it is specifically designed for document analysis and extraction of structured data such as parties, dates, and monetary amounts from PDFs and images. Its prebuilt models (e.g., the 'prebuilt-invoice' or 'prebuilt-contract' model) use optical character recognition (OCR) and deep learning to extract key-value pairs and entities from contracts without requiring custom training.

Answer

Azure AI Language (custom NER)

Answer

Azure AI Search with integrated vectorization

Answer

Azure OpenAI Service with GPT-4o

Question 2

You are designing a knowledge mining solution that must extract entities from scanned handwritten forms. The forms contain signatures and checkboxes. Which combination of Azure AI services should you recommend?

Accepted Answer

Azure AI Document Intelligence with a custom neural model and Azure AI Language for entity linking. The scenario requires extracting entities from scanned handwritten forms, which demands a custom neural model in Azure AI Document Intelligence to handle the variability of handwriting, signatures, and checkboxes. Azure AI Language's entity linking then enriches the extracted entities by disambiguating and linking them to a knowledge base, providing structured, meaningful output. The premade models in Document Intelligence are designed for printed text and common layouts, not handwritten content, making a custom neural model essential.

Answer

Azure AI Document Intelligence with a premade model and Azure AI Computer Vision

Answer

Azure AI Computer Vision (OCR) and Azure AI Search with integrated vectorization

Answer

Azure Cognitive Search and Azure AI Document Intelligence with a premade model

Question 3

Your organization has a knowledge base of technical manuals in PDF format. You need to enable users to ask natural language questions and get answers from the manuals. Which solution should you build?

Accepted Answer

Azure AI Language custom question answering with the documents as sources. Azure AI Language custom question answering is specifically designed to ingest documents (like PDFs) and provide a natural language Q&A interface over them. It uses a built-in extractive reader to find answer spans directly from the source text, making it the most straightforward solution for answering questions from a static knowledge base of technical manuals without requiring additional search or vectorization infrastructure.

Answer

Azure AI Search with integrated vectorization and semantic search

Answer

Azure OpenAI Service with GPT-4o and Azure AI Search as a data source

Answer

Azure AI Document Intelligence to extract text and then use Azure AI Search

Question 4

Your knowledge mining pipeline uses Azure AI Search with a custom skillset that calls an Azure Function. The function sometimes times out for large documents. What is the best way to handle this?

Accepted Answer

Increase the function timeout and ensure the function is in the same region as the search service. Increasing the function timeout (up to the Azure Functions maximum of 10 minutes for the Consumption plan or 30 minutes for the Premium plan) and ensuring the function is in the same region as the search service reduces latency and network overhead, directly addressing timeout issues for large documents. Azure AI Search custom skills must complete within the function's configured timeout, and regional colocation minimizes cross-region data transfer delays.

Answer

Use Azure AI Document Intelligence instead of a custom skill

Answer

Set the function timeout to the maximum of 24 hours

Answer

Move the custom skill to Azure AI Language custom entity recognition

Question 5

Which TWO actions should you take to ensure that an Azure AI Search indexer can access data from an Azure Storage account that contains sensitive data?

Accepted Answer

Configure the search service to use a system-assigned managed identity. Option C is correct because configuring the search service with a system-assigned managed identity lets the indexer authenticate to Azure Storage via Microsoft Entra ID (Azure RBAC role such as Storage Blob Data Reader) without embedding secrets, which is the recommended secure access pattern for sensitive data. Option D is correct because when the storage account firewall is enabled, the indexer's outbound requests come from the search service's IP address, so adding that IP to the storage account's network rules (via the 'Allow access from' selected networks list) permits the indexer to reach the data while keeping the firewall in place. Option A is not required here because a private endpoint is an alternative network-isolation approach, not a necessary action for enabling indexer access, and it is not one of the marked correct answers. Option B is not ideal for sensitive data because SAS keys are shared secrets that can be leaked or over-scoped, and managed identity is preferred. Option E is wrong because disabling the storage account firewall removes a critical security control and exposes the sensitive data, contradicting the goal of secure access.

Answer

Create a private endpoint in the storage account for the search service

Answer

Use a shared access key (SAS) in the data source definition

Answer

Disable the storage account firewall entirely

Question 6

Which THREE components are required to build a knowledge mining solution using Azure AI Search that extracts and enriches content from PDF files?

Accepted Answer

Skillset. In an Azure AI Search knowledge mining pipeline, the Data source (D) defines the connection to the source content (e.g., Azure Blob Storage holding the PDFs) and is required so the service knows what to crawl. The Indexer (E) is required because it drives the execution of the pipeline, reading from the data source, invoking the skillset, and writing enriched output into the index. The Skillset (B) is required to perform the AI enrichment (e.g., OCR, key phrase extraction, entity recognition) that extracts and enriches content from the PDFs. An Index (A) is also part of a full solution, but it is not one of the three components the question asks for in this specific set, and a Semantic configuration (C) is an optional enhancement for ranking/relevance, not a requirement for extraction and enrichment.

Answer

Index

Answer

Semantic configuration

Question 7

You are reviewing a skillset definition for an Azure AI Search indexer. The indexer is configured to index 1000 PDF documents. After running the indexer, you notice that only 500 documents have sentiment scores. What is the most likely cause?

Accepted Answer

The SplitSkill outputs are not correctly mapped to the SentimentSkill input; the skill runs only on the first page of each document. The SplitSkill divides each PDF into pages, and by default the SentimentSkill runs in the context of each page (e.g., /document/pages/*). If the SplitSkill outputs are not correctly mapped to the SentimentSkill input, the sentiment skill may only process the first page or fail to receive the split content, resulting in sentiment scores for only a subset of documents. Correct mapping ensures that each page's content is passed to the SentimentSkill for analysis.

Answer

The skills are defined in the wrong order; sentiment should run before split

Answer

The SentimentSkill is not supported in this region

Answer

The context of the SentimentSkill should be "/document" instead of "/document/pages/*"

Question 8

You review the configuration for an Azure AI Search indexer. The indexer runs successfully but no documents are indexed. What is the most likely cause?

Accepted Answer

The indexer does not have a skillset or field mappings to populate the 'content' field from the blob. When an Azure AI Search indexer runs successfully but indexes zero documents, the most common cause is that the indexer lacks a skillset or field mappings to extract and map the blob content (e.g., the 'content' field) into the search index. Without explicit field mappings or a skillset that defines how to populate the index fields from the blob's default properties, the indexer may complete execution without transferring any data into the index fields.

Answer

The connection string in the data source is invalid

Answer

The container name is incorrect

Answer

The maxFailedItems parameter is set to -1, which causes the indexer to skip all documents

Question 9

You are using Azure AI Search to index a set of PDF documents. The index includes a 'content' field with the extracted text. Users report that when they search for 'budget forecast', documents containing only 'budget' or 'forecast' are ranked lower than expected. Which configuration change would improve the ranking for multi-word queries?

Accepted Answer

Enable semantic search on the index. Semantic search in Azure AI Search uses deep neural network models to understand the intent and context of multi-word queries like 'budget forecast', rather than relying solely on keyword matching. This improves ranking by capturing the semantic relationship between terms, so documents containing both words (or conceptually related content) are ranked higher even if the exact phrase is not present.

Answer

Add a separate field for each word in the document

Answer

Change the analyzer to a custom analyzer that splits on spaces only

Answer

Set the 'content' field to a higher boosting value

Question 10

You are building an Azure AI Search solution that indexes data from multiple sources, including SQL Database and Azure Blob Storage. The index must be updated within 15 minutes of any source change. Which approach should you use to achieve near-real-time indexing?

Accepted Answer

Use the push API to send updates as soon as data changes. The push API (Azure Cognitive Search REST API or SDK) allows you to directly upload documents to the index as soon as data changes occur, bypassing the indexer's polling cycle. This provides sub-minute latency, meeting the 15-minute near-real-time requirement. Indexers with schedules or enrichment pipelines introduce inherent delays and are not designed for sub-minute updates.

Answer

Enable incremental enrichment on the skillset

Answer

Use an indexer with a schedule set to run every 5 minutes

Answer

Enable semantic search to speed up indexing

Question 11

You are building a solution to extract customer feedback from PDF documents stored in Azure Blob Storage. The solution must extract key phrases and sentiment scores, but you cannot use any pre-built models from Azure AI Language. What should you use?

Accepted Answer

Train custom models in Azure AI Language for key phrase extraction and sentiment analysis. Azure AI Language supports training custom models for named entity recognition (NER) to extract key phrases and custom text classification to predict sentiment labels. You can first extract text from PDFs using Azure AI Document Intelligence or other OCR services, then pass the text to the custom Language models. This approach meets the requirement without using any pre-built Azure AI Language models. Options A and C rely on pre-built APIs and are disallowed. Option D only provides OCR and cannot perform semantic analysis.

Answer

Use the sentiment analysis capability in Azure AI Language

Answer

Use Azure AI Language's pre-built key phrase extraction API

Answer

Use Azure AI Document Intelligence with the pre-built read model

Question 12

Your company has a large collection of legal contracts in PDF format stored in Azure Blob Storage. You need to extract key clauses, parties, and effective dates using a custom model in Azure AI Document Intelligence. The model must be retrained monthly as new contract templates are added. What is the recommended approach to handle model versioning and retraining?

Accepted Answer

Train a new model version using the 'compose' operation or copy the existing model and retrain with new samples. Azure AI Document Intelligence supports model versioning through the 'compose' operation, which allows you to combine multiple trained models into a single composed model, or by copying an existing model and retraining it with new samples. This approach preserves the existing model's knowledge while incrementally updating it with new contract templates, avoiding the need to retrain from scratch each month.

Answer

Use a multi-model ensemble by training separate models per template

Answer

Retrain the model from scratch each month using all historical data

Answer

Use Azure Machine Learning pipelines to automate retraining and deploy a new endpoint

Question 13

You need to extract personally identifiable information (PII) from a set of text documents before indexing them in Azure AI Search. The PII must be redacted. Which Azure AI service and configuration should you use?

Accepted Answer

Use the built-in PII detection skill in Azure AI Search with redaction mode enabled. Azure AI Search provides a built-in PII detection skill that can automatically identify and redact PII entities (such as names, phone numbers, and email addresses) from text documents during the indexing pipeline. The skill supports a 'redactionMode' configuration that replaces detected PII with a placeholder (e.g., '***'), meeting the requirement to redact PII before indexing.

Answer

Use the entity recognition skill in Azure AI Search and map to a target field

Answer

Use Azure AI Document Intelligence with a custom model to identify PII fields

Answer

Use Azure AI Language's key phrase extraction to find PII

Question 14

Which THREE components are essential when building a custom skill for Azure AI Search?

Accepted Answer

A Web API endpoint that processes documents. A custom skill in Azure AI Search is fundamentally a Web API endpoint (option A) that accepts a JSON payload of documents and returns enriched data, so this endpoint is essential. Option B is correct because field mappings (specifically input/output field mappings in the skillset) are required to pass data between the skill and the indexer, wiring the skill's outputs into the enrichment tree and ultimately the index. Option E is correct because the skill must define its inputs and outputs in JSON format within the skillset definition, specifying the @odata.type, context, and input/output mappings so the indexer knows what to send and receive. Option C is not required: a custom skill can call any HTTP-accessible logic, not necessarily an Azure Machine Learning-trained model. Option D is not required: Azure Functions are one possible host for the Web API, but the skill is invoked by the indexer during enrichment, not triggered on a schedule.

Answer

A machine learning model trained in Azure Machine Learning

Answer

An Azure Function to trigger the skill on a schedule

Question 15

Which TWO Azure AI services can be used to extract text from images as part of an Azure AI Search enrichment pipeline?

Accepted Answer

Azure AI Document Intelligence's read model. Option D is correct because Azure AI Document Intelligence's read model is designed to extract printed and handwritten text from documents and images, and it can be invoked as a custom skill within an Azure AI Search enrichment pipeline to produce text for downstream indexing. Option E is correct because Azure AI Search provides a built-in OCR skill (part of the cognitive skillset) that extracts text from image files in the enrichment pipeline, enabling text extraction directly without an external service call. Option A is incorrect because Azure Bot Service is for building conversational bots, not for extracting text from images. Option B is incorrect because Azure AI Speech to text transcribes spoken audio into text, not text from images. Option C is incorrect because Azure AI Language translation translates text between languages; it does not perform optical character recognition on images.

Answer

Azure Bot Service

Answer

Azure AI Speech to text

Answer

Azure AI Language translation

Question 16

You are reviewing the skillset definition for an Azure AI Search indexer. The SplitSkill splits the document content into pages of 5000 characters. The SentimentSkill is set to run on each page. However, the sentiment analysis is not producing correct results. What is the most likely cause?

Accepted Answer

The context of the SentimentSkill is set to an array, which is not supported. Setting the context of the SentimentSkill to an array (e.g., '/document/pages') causes the skill to run once on the entire array of pages, rather than iterating over each page individually. This results in incorrect sentiment analysis because the skill receives a collection of strings as a single input, which is not the intended per-page processing. The correct configuration is to set the context to '/document/pages/*' so that the skill executes on each page separately.

Answer

The maximumPageLength of 5000 is too high for sentiment analysis

Answer

The input source for SentimentSkill should be '/document/pages/*' but the SplitSkill output is named 'pages', so the input should be '/document/pages'

Answer

The SentimentSkill uses an incorrect @odata.type version

Question 17

You have an Azure AI Search indexer that is configured to index PDF files from Azure Blob Storage. The indexer is not extracting any text from the PDFs, and no errors are reported. You review the indexer definition as shown. What is the most likely cause?

Accepted Answer

The parsingMode is set to 'json' instead of 'default' or 'text'. The parsingMode set to 'json' tells the indexer to expect JSON files, not PDFs. Since PDFs are binary or text-based, the indexer cannot extract any content, but because it does not fail (JSON parsing simply returns no text), no error is reported. Changing parsingMode to 'default' or 'text' allows the indexer to correctly parse the PDF binary stream and extract text.

Answer

The field mapping from 'content' to 'content' is redundant and causes a conflict

Answer

The field mapping for 'metadata_storage_path' should be to 'metadata_storage_path'

Answer

The dataToExtract is set to 'contentAndMetadata' which is not supported for PDFs

Question 18

You executed the Azure CLI command shown to create an indexer. However, the indexer fails to run. The error indicates that the data source connection string is invalid. You have verified that the connection string is correct. What is the most likely issue?

Accepted Answer

The data source's connection string is not properly stored or the data source is not configured with the correct credentials. The error message explicitly states that the data source connection string is invalid, even though you have verified the string itself is correct. This typically occurs when the connection string is not stored correctly in the search service (e.g., it was truncated, escaped improperly, or contains special characters that were misinterpreted) or when the data source configuration lacks the correct credentials (e.g., the managed identity is not enabled, or the key/secret is missing). Option D directly addresses this mismatch between a verified string and the service's stored configuration.

Answer

The --query parameter is incorrectly formatted

Answer

The skillset name does not exist in the search service

Answer

The parsingMode should be 'json' for PDF files

Question 19

You are designing a knowledge mining solution for a large legal firm. The solution must extract key clauses, parties, and dates from thousands of PDF contracts. You need to minimize manual labeling effort while achieving high extraction accuracy. Which Azure AI service should you use?

Accepted Answer

Azure AI Document Intelligence custom extraction model. Azure AI Document Intelligence custom extraction model is the correct choice because it is specifically designed to extract structured fields (like clauses, parties, and dates) from documents such as PDF contracts. It uses a prebuilt layout model combined with custom training on a small set of labeled documents, minimizing manual labeling effort while achieving high accuracy through transfer learning and table/key-value pair extraction.

Answer

Azure OpenAI Service with GPT-4 prompt engineering

Answer

Azure AI Search with built-in blob indexing

Answer

Azure AI Language custom named entity recognition

Question 20

Your company uses Azure AI Search for an internal knowledge base. Users complain that searches for 'annual report 2023' return irrelevant results. You analyze the search index and find that the content field contains large blocks of text from PDFs. You need to improve relevance without re-indexing all documents. Which approach should you take?

Accepted Answer

Enable semantic ranking on the search index. Semantic ranking re-ranks search results using deep learning models to understand the intent and context of the query, rather than just keyword matching. Since the content field contains large text blocks from PDFs, semantic ranking can extract the most relevant passages and improve result relevance without requiring re-indexing or modifying the existing index schema.

Answer

Enable spell correction in the search query

Answer

Add a custom scoring profile based on term frequency

Answer

Change the index analyzer to a different language