Extracting Searchable Video Metadata with AWS AI Services

AIP-C01 By islever Editorial Last updated

Implement data validation and processing pipelines for FM consumption. Implement application integration patterns and development tools.

Short answer

C Create a SageMaker AI pipeline orchestrating Transcribe, Rekognition, and Comprehend to extract unified metadata and store it in DynamoDB with time-based partitioning.

Question

A media company needs to automatically generate searchable metadata for a video library that contains 10,000 hours of content. Each video includes visual scenes, spoken dialogue, background music, and on-screen text.
The company needs a solution to extract scene descriptions, speaker transcripts with timestamps, text overlay content, and emotional context from each video. The output must synchronize all metadata types with one another and create a searchable index. Users must be able to search based on any metadata type and jump directly to a specific moment in each video.
What solution will meet these requirements?
  1. Create an Amazon SageMaker AI processing job that extracts video frames at regular intervals. Use Amazon Transcribe to process audio and perform speech-to-text conversions. Use an Anthropic Claude model in Amazon Bedrock to process frames individually to produce multimodal scene analyses. Store outputs in separate Amazon S3 buckets that are organized by content type.
  2. Create an Amazon SageMaker AI processing job that extracts video frames at fixed intervals and attaches timestamps. Use Amazon Transcribe with time markers enabled to process audio. Use an Anthropic Claude model in Amazon Bedrock to analyze frame-transcript pairs. Store outputs in a unified JSON structure that preserves temporal relationships.
  3. Create an Amazon SageMaker AI parallel processing job that is orchestrated by a SageMaker AI pipeline. Configure multiple steps in the pipeline to extract video segments, process audio streams by using Amazon Transcribe, analyze video frames by using Amazon Rekognition to detect text in scenes, use Amazon Comprehend to perform sentiment analysis, and store unified metadata in Amazon DynamoDB with time-based partitioning. Correct Answer
  4. Create an Amazon SageMaker AI processing job that configures Anthropic Claude Opus in Amazon Bedrock to produce a comprehensive multimodal analysis for each video. Configure the processing job to extract metadata by using a single inference request and to index results in Amazon OpenSearch Service to provide temporal search capabilities.

Key Takeaways

Choosing specialized AWS AI services for specific extraction tasks (e.g., Transcribe for timestamps) is critical, as LLMs cannot reliably produce precise transcripts or timestamps in a single inference request.

This scenario requires extracting synchronized, multi-modal metadata from a video library to create a searchable index. The optimal solution uses specialized AWS AI services like Amazon Transcribe, Rekognition, and Comprehend within a pipeline, storing unified metadata in Amazon DynamoDB with time-based partitioning.

Common mistake: Choosing OpenSearch (Option D) for the search capability while ignoring the technical impossibility of extracting precise timestamps and transcripts from a long video via a single LLM inference request.

Expert Analysis

Why the Answer Is Correct

Option C correctly employs a pipeline of specialized AWS AI services tailored for each extraction requirement: Amazon Transcribe for speech-to-text with timestamps, Amazon Rekognition for detecting text overlays and scene elements, and Amazon Comprehend for emotional context and sentiment analysis. Storing the unified metadata in Amazon DynamoDB with time-based partitioning creates a searchable index that preserves temporal relationships, enabling users to jump directly to specific moments using timestamps as the sort key.

Why the Other Options Are Wrong

Option A stores outputs in separate S3 buckets, which completely fails the requirement to synchronize all metadata types with one another and lacks a searchable index. Option B creates a unified JSON structure but merely stores it in S3, which does not provide the searchable index required to query by any metadata type and jump to a moment. Option D uses Amazon OpenSearch Service for search but relies on a single Anthropic Claude Opus inference request to extract all metadata; LLMs cannot reliably perform speech-to-text with precise timestamps or process extremely long videos in a single request.

Community Comment Notes

As TNT87 noted, Option C leverages "highly specialized, native AWS AI services optimized for each precise requirement," which is the architecturally sound approach for guaranteed accuracy in tasks like timestamp generation and text detection compared to a generalized LLM approach.

Official Reference

https://docs.aws.amazon.com/transcribe/latest/dg/how-it-works.html

This link opens the official vendor documentation in a new tab.

Exam Strategy

When a question requires precise data extraction like timestamps or OCR, default to specialized AWS AI services (Transcribe, Rekognition) rather than foundational models. Eliminate options that store data in isolated silos when synchronization is required.

Community Votes

C
61%
B
28%
D
11%

61% of anonymous learners picked answer C. Votes are pick records left by other test-takers — they are not the verified answer.

Discussion 1 contributions

Compiled from public community threads and reader submissions, lightly edited for clarity. Handles and vote tallies are shown as recorded, and comments reflect the views of individual contributors.

From public community threads 1

  1. TNT87 👍 1 Answered C
    • Comprehensive Metadata Extraction: It leverages highly specialized, native AWS AI services optimized for each precise requirement: • Amazon Transcribe handles spoken dialogue and generates timestamps/time-markers for audio. • Amazon Rekognition natively detects text-in-video / text overlay content within the visual scenes. • Amazon Comprehend processes the resulting text to perform sentiment analysis, delivering the required emotional context.

Reader comments & corrections

No reader corrections yet — spotted an error or have a note? Share it below.

Log in to comment, report an error, or add a note about this question.

How Can AgentCore Query Cross-Account DynamoDB While Preserving User Identity?

S-grade

Which API Gateway setup streams Bedrock responses with least operational overhead?

S-grade

Resilient Retry and Streaming Strategies for Amazon Bedrock

S-grade

Bedrock Prompt Management & Guardrails Detection

S-grade

Practice All AIP-C01 Questions

Access 110 questions with complete answers and detailed explanations.

View Full AIP-C01 Practice Test →

← Back to AIP-C01 Study Guide