Applications of Foundation Models
Amazon Bedrock Core Capabilities
Amazon Bedrock is AWS managed access to foundation models and related generative AI tooling. Know the roles of Knowledge Bases, Agents, Guardrails, evaluation, customization, Prompt Management, Flows, logging, and throughput options.
Concepts
- Amazon Bedrock provides serverless API access to foundation models from AWS and third-party providers.
- Bedrock Knowledge Bases implement managed RAG over private data sources and vector stores.
- Bedrock Agents orchestrate multi-step tasks and call action groups such as Lambda functions or APIs.
- Bedrock Guardrails apply content filters, denied topics, sensitive information filters, and safety policies.
- Bedrock Model Evaluation compares model outputs with automatic and human evaluation workflows.
- Bedrock supports fine-tuning, continued pre-training for selected models, Custom Model Import, Provisioned Throughput, invocation logging, Prompt Management, and Flows.
Exam tips
- Knowledge Bases are the managed Bedrock RAG feature.
- Agents orchestrate tasks and can call action groups such as Lambda functions or APIs.
- Guardrails apply safety controls such as content filters, denied topics, and sensitive information filters.