Know Your Bots
This type of machine learning trains a model using labeled input-output pairs to make predictions.
What is supervised learning?
This AWS service converts spoken audio into written text.
What is Amazon Transcribe?
This SageMaker feature provides a fully managed, web-based IDE with Jupyter notebooks for ML development.
What is SageMaker Studio?
AWS defines six official responsible AI principles. Name any three of them.
What are any three of: fairness, explainability, privacy, robustness, governance, and transparency?
This AWS service provides serverless access to foundation models from Anthropic, AI21, Cohere, Stability AI, and Amazon through a single API.
What is Amazon Bedrock?
This is the phase where a trained ML model generates predictions on new, unseen data.
What is inference?
This AWS service uses NLP to extract entities, key phrases, and sentiment from unstructured text.
What is Amazon Comprehend?
This SageMaker capability automatically explores algorithms and hyperparameters to find the best model for your data with no ML expertise required."
What is SageMaker Autopilot?
This AWS service is used to build human-in-the-loop workflows that route low-confidence model predictions to a human reviewer.
What is Amazon Augmented AI (A2I)?
This technique customizes a foundation model for a specific task by continuing to train it on a smaller, labeled dataset, updating its weights.
What is fine-tuning?
A model that performs well on training data but poorly on new data is suffering from this problem.
What is overfitting?"
This AWS service is used to build conversational chatbots using voice and text.
What is Amazon Lex?
This SageMaker component is a purpose-built repository to store, share, and reuse ML features across teams and models.
What is SageMaker Feature Store?
This SageMaker feature creates structured documents capturing a model's intended use, training data, evaluation metrics, ethical considerations, and known limitations.
What are SageMaker Model Cards?
This GenAI technique improves response accuracy by retrieving relevant documents from a knowledge base and passing them as context to the model at inference time.
What is RAG (Retrieval-Augmented Generation)?
This ML technique reuses a model trained on a large dataset as the starting point for a new, related task.
What is transfer learning?
This AWS service goes beyond OCR to extract structured data such as tables and forms from scanned documents.
What is Amazon Textract?
This SageMaker tool detects bias in training data and models, and provides SHAP-based explainability for model predictions.
What is SageMaker Clarify?
This Amazon Titan feature embeds an invisible marker in AI-generated images to support transparency and help prevent misinformation.
What is Amazon Titan image watermarking?
This Amazon Bedrock feature enables a model to autonomously plan and execute multi-step tasks by invoking APIs and external tools.
What are Bedrock Agents?
This evaluation metric measures the fraction of actual positives correctly identified by a model, calculated as TP / (TP + FN).
What is recall (or sensitivity)?
This AWS service delivers real-time personalized product recommendations using the same ML technology as Amazon.com.
What is Amazon Personalize?
This SageMaker service compiles trained models to run up to 2x faster on specific target hardware including edge devices.
What is SageMaker Neo?
A Model Card can automatically pull bias and explainability reports from this SageMaker service, rather than requiring developers to enter them manually.
What is Amazon SageMaker Clarify?
This Amazon Bedrock feature lets developers configure content filters, topic denials, and PII redaction to control model inputs and outputs.
What are Bedrock Guardrails?