AWS Certified AI Practitioner (AIF-C01) glossary
Terms selected for AWS Certified AI Practitioner (AIF-C01) based on common objective language and practice focus.
Generative AI
AI capability that creates new content such as text, images, code, or audio based on prompts and learned patterns.
Read full term ->Foundation Model (FM)
Large pre-trained model that can be adapted or prompted for many downstream tasks.
Read full term ->Responsible AI
Framework for developing and deploying AI systems that are transparent, fair, accountable, and respect privacy and human oversight.
Read full term ->AI Risk Management
AI Risk Management involves identifying, assessing, and mitigating potential risks associated with AI systems, such as bias and misuse.
Read full term ->AI Security Fundamentals
AI Security Fundamentals cover the basic principles of safeguarding AI workloads, including access control and data protection.
Read full term ->Amazon Bedrock
Fully managed service that provides API access to foundation models from multiple providers.
Read full term ->Amazon SageMaker
AWS platform for building, training, tuning, and deploying machine learning models.
Read full term ->Artificial Intelligence (AI)
Broad field of building systems that perform tasks requiring human-like intelligence, such as reasoning, language understanding, and decision support.
Read full term ->Bias and Fairness
Concern that models may produce systematically skewed outcomes across groups due to data or design issues.
Read full term ->Content Grounding
Content Grounding refers to the practice of anchoring generative AI outputs in factual and reliable information to enhance accuracy.
Read full term ->Data Protection Compliance
Data Protection Compliance ensures that AI systems adhere to legal and regulatory standards for handling and processing data.
Read full term ->Deep Learning
Deep Learning is a subset of machine learning that utilizes neural networks with many layers to model complex patterns in data.
Read full term ->Embedding
Numerical representation of text or other data that captures semantic meaning for search and similarity tasks.
Read full term ->Fine-Tuning
Additional model training on task-specific data to improve behavior for a target domain.
Read full term ->Foundation Model Customization
Foundation Model Customization involves tailoring pre-trained models through techniques like prompting and fine-tuning to meet specific application needs.
Read full term ->Generative AI Use Cases
Generative AI Use Cases are practical applications where generative models are employed to create content, such as text, images, or music.
Read full term ->Hallucination
When a model generates fluent but incorrect or unsupported content.
Read full term ->Inference
Inference is the process of using a trained machine learning model to make predictions on new, unseen data.
Read full term ->Integration Patterns
Integration Patterns are strategies for embedding AI models into existing business workflows and systems to enhance functionality.
Read full term ->Machine Learning (ML)
Subset of AI where systems learn patterns from data to make predictions or decisions without explicit rule coding for every case.
Read full term ->Model Evaluation
Assessing the performance of a machine learning model using metrics such as accuracy, precision, and recall.
Read full term ->Guardrails
Controls that constrain model behavior, such as topic restrictions, harmful-content filters, and output policies.
Read full term ->Inference
Process of running a trained model to generate predictions or outputs for new inputs.
Read full term ->Model Training
The process of feeding data into a machine learning algorithm to learn the patterns and make predictions.
Read full term ->Model Types
Model Types refer to the various architectures and algorithms used in AI/ML, such as decision trees, neural networks, and support vector machines.
Read full term ->Operational Controls for AI
Operational Controls for AI are processes and technologies that ensure AI systems operate reliably and securely within an organization.
Read full term ->Overfitting
When a model learns training data too closely and performs poorly on new data.
Read full term ->Prompt Engineering
Crafting and refining input instructions to guide a generative model toward desired outputs.
Read full term ->Prompt Foundations
Prompt Foundations involve the principles and techniques used to effectively guide generative AI models in producing desired outputs.
Read full term ->Responsible AI Guidelines
Responsible AI Guidelines are principles and practices aimed at ensuring AI systems are developed and used ethically, respecting fairness, transparency, and privacy.
Read full term ->Retrieval-Augmented Generation (RAG)
Pattern that retrieves trusted context at query time and injects it into prompts to ground model responses.
Read full term ->Vector Database
Storage optimized for high-dimensional vectors and similarity search.
Read full term ->