AWS solutions documentation change
Summary
Added new documentation page for Generative AI Application Builder on AWS solution, describing its architecture, use cases (Agent Builder, Multi-Agent, MCP Server, RAG Chatbot, Management Dashboard), benefits, and deployment options.
Security assessment
The documentation describes security features like AWS WAF protection, Amazon Cognito authentication, custom Lambda authorizers for IAM policies, and built-in AWS Well-Architected security principles. However, there is no evidence this change addresses a specific security vulnerability or incident; it's a new solution introduction with standard security features documentation.
Diff
diff --git a/solutions/generative-ai-application-builder-on-aws/index.md b/solutions/generative-ai-application-builder-on-aws/index.md index 8b1378917..37d8982a9 100644 --- a//solutions/generative-ai-application-builder-on-aws/index.md +++ b//solutions/generative-ai-application-builder-on-aws/index.md @@ -0,0 +1,335 @@ +# Generative AI Application Builder on AWS + +Rapidly develop and deploy production-ready Agentic AI applications + +### About this deployment + +Version + 4.1.11 +Released + 4/2026 +Author + AWS +Est. deployment time + 10 mins +Estimated cost + [See details](/solutions/latest/generative-ai-application-builder-on-aws/cost.html) + +[Open guide](/solutions/latest/generative-ai-application-builder-on-aws/solution-overview.html)[Deploy now](https://console.aws.amazon.com/cloudformation/home?region=us-east-1#/stacks/quickcreate?&templateURL=https://solutions-reference.s3.amazonaws.com/generative-ai-application-builder-on-aws/latest/generative-ai-application-builder-on-aws.template&redirectId=SolutionWeb) + +## Overview + +Building and managing agentic AI applications can be challenging for teams without deep AI infrastructure or orchestration expertise. Generative AI Application Builder (GAAB) on AWS eliminates that complexity by unifying all the key components of modern agentic systems into one integrated solution. GAAB brings together the Model Context Protocol (MCP), agents, and multi-agent orchestration with built-in authentication, permissioning, networking, and application management—so teams can focus on iterating and improving their applications instead of managing infrastructure. + +With GAAB, you can deploy local MCP servers as remote endpoints using AWS Lambda, container images, or OpenAPI and Smithy API specifications. The platform enables developers to create agents, assign MCP tools, and orchestrate complex behaviors through the Agent-as-a-Tool framework. Once an agentic application is ready, teams can collaborate instantly by accessing the application out of the box or integrating it with their own client through a single WebSocket connection. + +GAAB accelerates the path from prototype to production by simplifying the deployment, management, and scaling of secure, extensible agentic applications built on AWS. + +## Benefits + +Rapid experimentation + +You can experiment with different model, system prompt, and memory configurations for the same use case to compare outputs and performance directly from the GAAB management dashboard. Once you have completed your testing iterations, you can easily deploy the GAAB created applications with the default user interface or integrate the agentic application with your existing system. + +Configurability + +Centralize the deployment and management of Model Context Protocol (MCP) servers, AI Agents, and Multi-agent workflows while governing and decentralizing access to the deployed agentic applications across your teams. + +Production-ready + +Built with AWS Well-Architected design principles, this solution offers enterprise-grade security and scalability with high availability and low latency, ensuring seamless integration into your applications with high performance standards. + +Extensible modular architecture + +Extend this solution’s functionality by integrating your existing projects or natively connecting additional AWS services. GAAB is an open-source application, enabling you to use the included Strands and LangChain orchestration layers and Lambda functions to connect with the services of your choice. + +## How it works + +Build and Deploy Agents + +The Agent Builder use case enables you to quickly configure, deploy, and manage AI Agents from the Management Dashboard. Configure the Agent's model, tools, memory, system prompt, and integration pattern before deploying the AI Agent stack. + +[Open implementation guide](/solutions/latest/generative-ai-application-builder-on-aws/architecture-overview.html)  Step 1 + +Admin users deploy the use case using the Deployment Dashboard. Business users sign in to the use case UI. + +Step 2 + +Amazon CloudFront delivers the web UI which is hosted in an Amazon S3 bucket. + +Step 3 + +The web UI leverages a WebSocket integration built using Amazon API Gateway. The API Gateway is backed by a custom AWS Lambda authorizer function, which returns the appropriate AWS Identity and Access Management (IAM) policy based on the Amazon Cognito group the authenticating user belongs to. The policy is stored in Amazon DynamoDB. + +Step 4 + +Amazon Cognito authenticates users and backs both the CloudFront web UI and API Gateway. + +Step 5 + +Incoming requests from the business user are passed from API Gateway to an Amazon Simple Queue Service queue and then to the Lambda function. The queue enables the asynchronous operation of the API Gateway to Lambda integration. The queue passes connection information to the Lambda function which will then post results directly back to the API Gateway websocket connection to support long running inference calls. + +Step 6 + +The Lambda function retrieves the agent configuration from DynamoDB. + +Step 7 + +Using the user input and any relevant use case configurations, the Lambda function builds and sends a request payload to the agent, running on Amazon Bedrock AgentCore Runtime + +Step 8 + +The agent connects to associated MCP servers and registers the tools to the strands agent instance. The agent then autonomously selects and performs actions based on tool descriptions and task requirements. + +Step 9 + +When the response comes back from the AgentCore runtime, the Lambda function streams the response back through the API Gateway WebSocket to be consumed by the client application. + +Multi-Agent Applications + +The Multi-Agent use case enables you to quickly orchestrate multiple Agents to address complex tasks from the Management Dashboard. Configure the orchestration Agent's model, the AI Agents to use as tools, the orchestration prompt, and integration pattern before deploying the Multi-Agent stack. + +[Open implementation guide](/solutions/latest/generative-ai-application-builder-on-aws/architecture-overview.html)  Step 1 + +Admin users deploy the workflow using the Deployment Dashboard, selecting Agent Builder agents to include as specialized agents. + +Step 2 + +Amazon CloudFront delivers the web UI which is hosted in an Amazon S3 bucket. + +Step 3 + +The web UI leverages a WebSocket integration built using Amazon API Gateway. The API Gateway is backed by a custom AWS Lambda authorizer function, which returns the appropriate AWS Identity and Access Management (IAM) policy based on the Amazon Cognito group the authenticating user belongs to. The policy is stored in Amazon DynamoDB. + +Step 4 + +Amazon Cognito authenticates users and backs both the CloudFront web UI and API Gateway. + +Step 5 + +Incoming requests from the business user are passed from API Gateway to an Amazon Simple Queue Service queue and then to the Lambda function. The queue enables the asynchronous operation of the API Gateway to Lambda integration. + +Step 6 + +The Lambda function retrieves workflow configuration from DynamoDB, including the list of specialized Agent Builder agents. + +Step 7 + +Using the user input and workflow configuration, Lambda sends requests to the Amazon Bedrock AgentCore Runtime hosting the supervisor agent. + +Step 8 + +The supervisor agent creates local instances of all specialized Agent Builder agents within the AgentCore Runtime environment. These specialized agents are registered as tools using the Agents as Tools pattern. The supervisor then autonomously selects and delegates work to specialized agents based on agent descriptions and task requirements. + +Step 9 + +The supervisor agent aggregates results from specialized agents and formulates the final response, returning it to the Lambda to be streamed back to the client application through the API Gateway Websocket. + +Deploy Remote MCP Servers + +The MCP Server use case enables you to quickly deploy and manage MCP Servers from the Management Dashboard. You can connect to any existing MCP Server by providing its endpoint URL, or deploy and host your own MCP Server via an image, Lambda function, OpenAPI spec, or Smithy file, providing the flexibility to integrate external third-party services or self-hosted MCP Servers for use in Agent Builder and Multi-Agent Use cases. + +[Open implementation guide](/solutions/latest/generative-ai-application-builder-on-aws/architecture-overview.html)  Step 1 + +Admin users deploy the MCP Server use case using the Deployment Dashboard, selecting either Gateway or Runtime deployment method. + +Step 2 + +This action is authenticated with Amazon Cognito. + +Step 3 + +For the Gateway deployment, the solution creates an Amazon Bedrock AgentCore Gateway that transforms existing AWS Lambda functions, APIs, or external MCP servers into MCP-compliant tools. For the Runtime deployment, the solution deploys containerized MCP servers on AgentCore Runtime using provided Amazon Elastic Container Registry images. + +Step 4 + +Gateway deployments retrieve the necessary API/Lambda/Smithy schemas from their uploaded location in Amazon Simple Storage Service (Amazon S3), or connect directly to MCP Server URL endpoints. + +Step 5 + +Runtime deployments retrieve the containerized MCP server provided by the user from Amazon ECR. + +Step 6 + +The MCP Server is instrumented with an AgentCore Identity OAuth client. + +Step 7 + +The MCP Server makes the associated tools available at the /mcp endpoint for Agents to discover. + +Step 8 + +Amazon CloudWatch collects operational metrics and logs from MCP server deployments for monitoring and troubleshooting. + +Retrieval Augmented Generation (RAG) Chatbots + +The RAG Chatbot use case enables you to quickly deploy and manage a Retrieval Augmented Generation (RAG) enabled chatbot. Configure the chatbot’s model, RAG knowledgebase, and guardrails to implement safeguards and reduce hallucinations before deploying the chatbot stack. + +[Open implementation guide](https://docs.aws.amazon.com/solutions/latest/generative-ai-application-builder-on-aws/architecture-overview.html)  Step 1 + +Admin users deploy the use case using the Deployment Dashboard. Business users log in to the use case UI. + +Step 2 + +CloudFront delivers the web UI which is hosted in an S3 bucket. + +Step 3 + +The web UI leverages a WebSocket integration built using API Gateway. The API Gateway is backed by a custom Lambda authorizer function, which returns the appropriate AWS Identity and Access Management (IAM) policy based on the Amazon Cognito group the authenticating user belongs to. The policy is stored in DynamoDB. + +Step 4 + +Amazon Cognito authenticates users and backs both the CloudFront web UI and API Gateway. + +Step 5 + +Incoming requests from the business user are passed from API Gateway to an Amazon SQS queue and then to the LangChain Orchestrator. The LangChain Orchestrator is a collection of Lambda functions and layers that provide the business logic for fulfilling requests coming from the business user. The queue enables the asynchronous operation of the API Gateway to Lambda integration. The queue passes connection information to the Lambda functions which will then post results directly back to the API Gateway websocket connection to support long running inference calls. + +Step 6 + +The LangChain Orchestrator uses Amazon DynamoDB to get the configured LLM options and necessary session information (such as the chat history). + +Step 7 + +If the deployment has a knowledge base enabled, then the LangChain Orchestrator leverages Amazon Kendra or Knowledge Bases for Amazon Bedrock to run a search query to retrieve document excerpts. +