Robots powered by Gemini Robotics models can learn complex actions like preparing salads and even folding an origami fox.Read More
Introducing AWS MCP Servers for code assistants (Part 1)
We’re excited to announce the open source release of AWS MCP Servers for code assistants — a suite of specialized Model Context Protocol (MCP) servers that bring Amazon Web Services (AWS) best practices directly to your development workflow. Our specialized AWS MCP servers combine deep AWS knowledge with agentic AI capabilities to accelerate development across key areas. Each AWS MCP Server focuses on a specific domain of AWS best practices, working together to provide comprehensive guidance throughout your development journey.
This post is the first in a series covering AWS MCP Servers. In this post, we walk through how these specialized MCP servers can dramatically reduce your development time while incorporating security controls, cost optimizations, and AWS Well-Architected best practices into your code. Whether you’re an experienced AWS developer or just getting started with cloud development, you’ll discover how to use AI-powered coding assistants to tackle common challenges such as complex service configurations, infrastructure as code (IaC) implementation, and knowledge base integration. By the end of this post, you’ll understand how to start using AWS MCP Servers to transform your development workflow and deliver better solutions, faster.
If you want to get started right away, skip ahead to the section “From Concept to working code in minutes.”
AI is transforming how we build software, creating opportunities to dramatically accelerate development while improving code quality and consistency. Today’s AI assistants can understand complex requirements, generate production-ready code, and help developers navigate technical challenges in real time. This AI-driven approach is particularly valuable in cloud development, where developers need to orchestrate multiple services while maintaining security, scalability, and cost-efficiency.
Developers need code assistants that understand the nuances of AWS services and best practices. Specialized AI agents can address these needs by:
- Providing contextual guidance on AWS service selection and configuration
- Optimizing compliance with security best practices and regulatory requirements
- Promoting the most efficient utilization and cost-effective solutions
- Automating repetitive implementation tasks with AWS specific patterns
This approach means developers can focus on innovation while AI assistants handle the undifferentiated heavy lifting of coding. Whether you’re using Amazon Q, Amazon Bedrock, or other AI tools in your workflow, AWS MCP Servers complement and enhance these capabilities with deep AWS specific knowledge to help you build better solutions faster.
Model Context Protocol (MCP) is a standardized open protocol that enables seamless interaction between large language models (LLMs), data sources, and tools. This protocol allows AI assistants to use specialized tooling and to access domain-specific knowledge by extending the model’s capabilities beyond its built-in knowledge—all while keeping sensitive data local. Through MCP, general-purpose LLMs can now seamlessly access relevant knowledge beyond initial training data and be effectively steered towards desired outputs by incorporating specific context and best practices.
Accelerate building on AWS
What if your AI assistant could instantly access deep AWS knowledge, understanding every AWS service, best practice, and architectural pattern? With MCP, we can transform general-purpose LLMs into AWS specialists by connecting them to specialized knowledge servers. This opens up exciting new possibilities for accelerating cloud development while maintaining security and following best practices.
Build on AWS in a fraction of the time, with best practices automatically applied from the first line of code. Skip hours of documentation research and immediately access ready-to-use patterns for complex services such as Amazon Bedrock Knowledge Bases. Our MCP Servers will help you write well-architected code from the start, implement AWS services correctly the first time, and deploy solutions that are secure, observable, and cost-optimized by design. Transform how you build on AWS today.
- Enforce AWS best practices automatically – Write well-architected code from the start with built-in security controls, proper observability, and optimized resource configurations
- Cut research time dramatically – Stop spending hours reading documentation. Our MCP Servers provide contextually relevant guidance for implementing AWS services correctly, addressing common pitfalls automatically
- Access ready-to-use patterns instantly – Use pre-built AWS CDK constructs, Amazon Bedrock Agents schema generators, and Amazon Bedrock Knowledge Bases integration templates that follow AWS best practices from the start
- Optimize cost proactively – Prevent over-provisioning as you design your solution by getting cost-optimization recommendations and generating a comprehensive cost report to analyze your AWS spending before deployment
To turn this vision into reality and make AWS development faster, more secure, and more efficient, we’ve created AWS MCP Servers—a suite of specialized AWS MCP Servers that bring AWS best practices directly to your development workflow. Our specialized AWS MCP Servers combine deep AWS knowledge with AI capabilities to accelerate development across key areas. Each AWS MCP Server focuses on a specific domain of AWS best practices, working together to provide comprehensive guidance throughout your development journey.
Overview of domain-specific MCP Servers for AWS development
Our specialized MCP Servers are designed to cover distinct aspects of AWS development, each bringing deep knowledge to specific domains while working in concert to deliver comprehensive solutions:
- Core – The foundation server that provides AI processing pipeline capabilities and serves as a central coordinator. It helps provide clear plans for building AWS solutions and can federate to other MCP servers as needed.
- AWS Cloud Development Kit (AWS CDK) – Delivers AWS CDK knowledge with tools for implementing best practices, security configurations with cdk-nag, Powertools for AWS Lambda integration, and specialized constructs for generative AI services. It makes sure infrastructure as code (IaC) follows AWS Well-Architected principles from the start.
- Amazon Bedrock Knowledge Bases – Enables seamless access to Amazon Bedrock Knowledge Bases so developers can query enterprise knowledge with natural language, filter results by data source, and use reranking for improved relevance.
- Amazon Nova Canvas – Provides image generation capabilities using Amazon Nova Canvas through Amazon Bedrock, enabling the creation of visuals from text prompts and color palettes—perfect for mockups, diagrams, and UI design concepts.
- Cost – Analyzes AWS service costs and generates comprehensive cost reports, helping developers understand the financial implications of their architectural decisions and optimize for cost-efficiency.
Prerequisites
To complete the solution, you need to have the following prerequisites in place:
- uv package manager
- Install Python using
uv python install 3.13 - AWS credentials with appropriate permissions
- An MCP-compatible LLM client (such as Anthropic’s Claude for Desktop, Cline, Amazon Q CLI, or Cursor)
From concept to working code in minutes
You can download the AWS MCP Servers on GitHub or through the PyPI package manager. Here’s how to get started using your favorite code assistant with MCP support.
To install MCP Servers, enter the following code:
AWS MCP Servers in action
Here’s how AWS MCP servers transform the development experience:
Developer: “I need to build an AI-powered chatbot using Amazon Bedrock that can answer questions from our company’s knowledge base. I also want to add a tool for the chatbot to call our internal API.”
Core: “I’ll help you build an Amazon Bedrock Knowledge Bases chatbot with API integration. Let’s create an architecture that uses Amazon Bedrock Agents with a custom action group to call your internal API.”
Core generates a comprehensive architecture diagram showing the knowledge base integration, Amazon Bedrock Agents configuration with action groups, API connectivity, and data flow between components.
AWS CDK: “Here’s the infrastructure code for your chatbot with the Amazon Bedrock Agents action group. I’ve included proper IAM roles, security controls, and Lambda Powertools for observability.”
The CDK MCP Server generates complete AWS CDK code to deploy the entire solution. It automatically runs cdk-nag to identify potential security issues and provides remediation steps for each finding, making sure that the infrastructure follows AWS Well-Architected best practices.
Amazon Bedrock Knowledge Bases retrieval: “I’ve configured the optimal settings for your knowledge base queries, including proper reranking for improved relevance.”
Amazon Bedrock Knowledge Bases MCP Server demonstrates how to structure queries to the knowledge base for maximum relevance, provides sample code for filtering by data source, and shows how to integrate the knowledge base responses with the chatbot interface.
Amazon Nova Canvas: “To enhance your chatbot’s capabilities, I’ve created visualizations that can be generated on demand when users request data explanations.”
Amazon Nova Canvas MCP server generates sample images showing how Amazon Nova Canvas can create charts, diagrams, and visual explanations based on knowledge base content, making complex information more accessible to users.
Cost Analysis: “Based on your expected usage patterns, here’s the estimated monthly cost breakdown and optimization recommendations.”
The Cost Analysis MCP Server generates a detailed cost analysis report showing projected expenses for each AWS service, identifies cost optimization opportunities such as reserved capacity for Amazon Bedrock, and provides specific recommendations to reduce costs without impacting performance.
With AWS MCP Servers, what would typically take days of research and implementation is completed in minutes, with better quality, security, and cost-efficiency than manual development in that same time.
Best practices for MCP-assisted development
To maximize the benefits of MCP assisted development while maintaining security and code quality, developers should follow these essential guidelines:
- Always review generated code for security implications before deployment
- Use MCP Servers as accelerators, not replacements for developer judgment and expertise
- Keep MCP Servers updated with the latest AWS security best practices
- Follow the principle of least privilege when configuring AWS credentials
- Run security scanning tools on generated infrastructure code
Coming up in the series
This post introduced the foundations of AWS MCP Servers and how they accelerate AWS development through specialized, AWS specific MCP Servers. In upcoming posts, we’ll dive deeper into:
- Detailed walkthroughs of each MCP server’s capabilities
- Advanced patterns for integrating AWS MCP Servers into your development workflow
- Real-world case studies showing AWS MCP Servers’ impact on development velocity
- How to extend AWS MCP Servers with your own custom MCP servers
Stay tuned to learn how AWS MCP Servers can transform your specific AWS development scenarios and help you build better solutions faster. Visit our GitHub repository or Pypi package manager to explore example implementations and get started today.
About the Authors
Jimin Kim is a Prototyping Architect on the AWS Prototyping and Cloud Engineering (PACE) team, based in Los Angeles. With specialties in Generative AI and SaaS, she loves helping her customers succeed in their business. Outside of work, she cherishes moments with her wife and three adorable calico cats.
Pranjali Bhandari is part of the Prototyping and Cloud Engineering (PACE) team at AWS, based in the San Francisco Bay Area. She specializes in Generative AI, distributed systems, and cloud computing. Outside of work, she loves exploring diverse hiking trails, biking, and enjoying quality family time with her husband and son.
Laith Al-Saadoon is a Principal Prototyping Architect on the Prototyping and Cloud Engineering (PACE) team. He builds prototypes and solutions using generative AI, machine learning, data analytics, IoT & edge computing, and full-stack development to solve real-world customer challenges. In his personal time, Laith enjoys the outdoors–fishing, photography, drone flights, and hiking.
Paul Vincent is a Principal Prototyping Architect on the AWS Prototyping and Cloud Engineering (PACE) team. He works with AWS customers to bring their innovative ideas to life. Outside of work, he loves playing drums and piano, talking with others through Ham radio, all things home automation, and movie nights with the family.
Justin Lewis leads the Emerging Technology Accelerator at AWS. Justin and his team help customers build with emerging technologies like generative AI by providing open source software examples to inspire their own innovation. He lives in the San Francisco Bay Area with his wife and son.
Anita Lewis is a Technical Program Manager on the AWS Emerging Technology Accelerator team, based in Denver, CO. She specializes in helping customers accelerate their innovation journey with generative AI and emerging technologies. Outside of work, she enjoys competitive pickleball matches, perfecting her golf game, and discovering new travel destinations.
Harness the power of MCP servers with Amazon Bedrock Agents
AI agents extend large language models (LLMs) by interacting with external systems, executing complex workflows, and maintaining contextual awareness across operations. Amazon Bedrock Agents enables this functionality by orchestrating foundation models (FMs) with data sources, applications, and user inputs to complete goal-oriented tasks through API integration and knowledge base augmentation. However, in the past, connecting these agents to diverse enterprise systems has created development bottlenecks, with each integration requiring custom code and ongoing maintenance—a standardization challenge that slows the delivery of contextual AI assistance across an organization’s digital ecosystem. This is a problem that you can solve by using Model Context Protocol (MCP), which provides a standardized way for LLMs to connect to data sources and tools.
Today, MCP is providing agents standard access to an expanding list of accessible tools that you can use to accomplish a variety of tasks. In time, MCP can promote better discoverability of agents and tools through marketplaces, enabling agents to share context and have common workspaces for better interaction, and scale agent interoperability across the industry.
In this post, we show you how to build an Amazon Bedrock agent that uses MCP to access data sources to quickly build generative AI applications. Using Amazon Bedrock Agents, your agent can be assembled on the fly with MCP-based tools as in this example:
We showcase an example of building an agent to understand your Amazon Web Service (AWS) spend by connecting to AWS Cost Explorer, Amazon CloudWatch, and Perplexity AI through MCP. You can use the code referenced in this post to connect your agents to other MCP servers to address challenges for your business. We envision a world where agents have access to an ever-growing list of MCP servers that they can use for accomplishing a wide variety of tasks.
Model Context Protocol
Developed by Anthropic as an open protocol, MCP provides a standardized way to connect AI models to virtually any data source or tool. Using a client-server architecture, MCP enables developers to expose their data through lightweight MCP servers while building AI applications as MCP clients that connect to these servers. Through this architecture, MCP enables users to build more powerful, context-aware AI agents that can seamlessly access the information and tools they need. Whether you’re connecting to external systems or internal data stores or tools, you can now use MCP to interface with all of them in the same way. The client-server architecture of MCP enables your agent to access new capabilities as the MCP server updates without requiring any changes to the application code.
MCP architecture
MCP uses a client-server architecture that contains the following components and is shown in the following figure:
- Host: An MCP host is a program or AI tool that requires access to data through the MCP protocol, such as Claude Desktop, an integrated development environment (IDE), or any other AI application.
- Client: Protocol clients that maintain one-to-one connections with servers.
- Server: Lightweight programs that expose capabilities through standardized MCP.
- Local data sources: Your databases, local data sources, and services that MCP servers can securely access.
- Remote services: External systems available over the internet through APIs that MCP servers can connect to.
Let’s walk through how to set up Amazon Bedrock agents that take advantage of MCP servers.
Using MCP with Amazon Bedrock agents
In this post, we provide a step-by-step guide for how to connect your favorite MCP servers with Amazon Bedrock agents as Action Groups that an agent can use to accomplish tasks provided by the user. The AgentInlineSDK provides a straightforward way to create inline agents, containing a built-in MCP client implementation that provides you with direct access to tools delivered by an MCP server.
As part of creating an agent, the developer creates an MCP client specific to each MCP server that requires agent communication. When invoked, the agent determines which tools are needed for the user’s task; if MCP server tools are required, it uses the corresponding MCP client to request tool execution from that server. The user code doesn’t need to be aware of the MCP protocol because that’s handled by the MCP client provided the InlineAgent code repository.
To orchestrate this workflow, you take advantage of the return control capability of Amazon Bedrock Agents. The following diagram illustrates the end-to-end flow of an agent handling a request that uses two tools. In the first flow, a Lambda-based action is taken, and in the second, the agent uses an MCP server.
Use case: transform how you manage your AWS spend across different AWS services including Amazon Bedrock
To show how an Amazon Bedrock agent can use MCP servers, let’s walk through a sample use case. Imagine asking questions like “Help me understand my Bedrock spend over the last few weeks” or “What were my EC2 costs last month across regions and instance types?” and getting a human-readable analysis of the data instead of raw numbers on a dashboard. The system interprets your intent and delivers precisely what you need—whether that’s detailed breakdowns, trend analyses, visualizations, or cost-saving recommendations. This is useful because what you’re interested in is insights rather than data. You can accomplish this using two MCP servers: a custom-built MCP server for retrieving the AWS spend data and an open source MCP server from Perplexity AI to interpret the data. You add these two MCP servers as action groups in an inline Amazon Bedrock agent. This gives you an AI agent that can transform the way you manage your AWS spend. All the code for this post is available in the GitHub repository.
Let’s walk through how this agent is created using inline agents. You can use inline agents to define and configure Amazon Bedrock agents dynamically at runtime. They provide greater flexibility and control over agent capabilities, enabling users to specify FMs, instructions, action groups, guardrails, and knowledge bases as needed without relying on pre-configured control plane settings. It’s worth noting that you can also orchestrate this behavior without inline agents by using RETURN_CONTROL with the InvokeAgent API.
MCP components in Amazon Bedrock Agents
- Host: This is the Amazon Bedrock inline agent. This agent adds MCP clients as action groups that can be invoked through RETURN_CONTROL when the user asks an AWS spend-related question.
- Client: You create two clients that establish one-to-one connections with their respective servers: a cost explorer client with specific cost server parameters and a Perplexity AI client with Perplexity server parameters.
- Servers: You create two MCP servers that each run locally on your machine and communicate to your application over standard input/output (alternatively, you could also configure the client to talk to remote MCP servers).
- Cost Explorer and Amazon CloudWatch Logs (for Amazon Bedrock model invocation log data) and an MCP server to retrieve the AWS spend data.
- Perplexity AI MCP server to interpret the AWS spend data.
- Data sources: The MCP servers talk to remote data sources such as Cost Explorer API, CloudWatch Logs and the Perplexity AI search API.
Prerequisites
You need the following prerequisites to get started implementing the solution in this post:
- An AWS account
- Familiarity with FMs and Amazon Bedrock
- Install AWS Command Line Interface (AWS CLI) and set up credentials
- Python 3.11 or later
- AWS Cloud Development Kit (AWS CDK) CLI
- Enable model access for Anthropic’s Claude 3.5 Sonnet v2
- You need to have your
AWS_ACCESS_KEY_IDandAWS_SECRET_ACCESS_KEYso that you can set them using environment variables for the server - The two MCP servers are run as Docker daemons, so you need to have Docker installed and running on your computer
The MCP servers run locally on your computer and need to access AWS services and the Perplexity API. You can read more about AWS credentials in Manage access keys for IAM users. Make sure that your credentials include AWS Identity and Access Manager (IAM) read access to Cost Explorer and CloudWatch. You can do this by using AWSBillingReadOnlyAccess and CloudWatchReadOnlyAccess managed IAM permissions. You can get the Perplexity API key from the Perplexity Sonar API page.
Steps to run
With the prerequisites in place, you’re ready to implement the solution.
- Navigate to the InlineAgent GitHub repository.
- Follow the setup steps.
- Navigate to the cost_explorer_agent This folder contains the code for this post.
- Create a
.envfile incost_explorer_agentdirectory using example. - Build
aws-cost-explorer-mcpserver - You’re now ready to create an agent that can invoke these MCP servers to provide insights into your AWS spend. You can do this by running the python
main.pycommand. The output will look like the following example. The agent connects with the two MCP servers and accesses their respective tools. Next, the agent lays out a plan to use specific tools sequentially, uses code interpreter to generate a chart showing cost distribution, and uses Cost Explorer and Perplexity API to provide information about Amazon Bedrock and the AWS account’s cost expenditure.
Implementation details
Now that you understand the output produced by an agent, let’s lift the curtain and review some of the important pieces of code that produce the output.
- Creating MCP clients: config.py, defines the two MCP clients that talk to your two MCP servers.
- Server parameters are defined for the cost explorer and Perplexity clients. The solution uses
StdioServerParameters, which configures how the client should communicate over standard input/output (stdio) streams. This contains the parameters required by the server to access the required data through APIs. - In
main.py, the MCP server parameters are imported and used to create your two MCP clients.
- Server parameters are defined for the cost explorer and Perplexity clients. The solution uses
- Configure agent action group:
main.pycreates the action group that combines the MCP clients into a single interface that the agent can access. This enables the agent to ask your application to invoke either of these MCP servers as needed through return of control. - Inline agent creation: The inline agent can be created with the following specifications:
- Foundation model: Configure your choice of FM to power your agent. This can be any model provided on Amazon Bedrock. This example uses Anthropic’s Claude 3.5 Sonnet model.
- Agent instruction: Provide instructions to your agent that contain the guidance and steps for orchestrating responses to user queries. These instructions anchor the agent’s approach to handling various types of queries
- Agent name: Name of your agent.
- Action groups: Define the action groups that your agent can access. These can include single or multiple action groups, with each group having access to multiple MCP clients or AWS Lambda As an option, you can configure your agent to use Code Interpreter to generate, run, and test code for your application.
You can use this example to build an inline agent on Amazon Bedrock that establishes connections with different MCP servers and groups their clients into a single action group for the agent to access.
Conclusion
The Anthropic MCP protocol offers a standardized way of connecting FMs to data sources, and now you can use this capability with Amazon Bedrock Agents. In this post, you saw an example of combining the power of Amazon Bedrock and MCP to build an application that offers a new perspective on understanding and managing your AWS spend.
Organizations can now offer their teams natural, conversational access to complex financial data while enhancing responses with contextual intelligence from sources like Perplexity. As AI continues to evolve, the ability to securely connect models to your organization’s critical systems will become increasingly valuable. Whether you’re looking to transform customer service, streamline operations, or gain deeper business insights, the Amazon Bedrock and MCP integration provides a flexible foundation for your next AI innovation. You can dive deeper on this MCP integration by exploring our code samples.
Here are some examples of what you can build by connecting your Amazon Bedrock Agents to MCP servers:
- A multi-data source agent that retrieves data from different data sources such as Amazon Bedrock Knowledge Bases, Sqlite, or even your local filesystem.
- A developer productivity assistant agent that integrates with Slack and GitHub MCP servers.
- A machine learning experiment tracking agent that integrates with the Opik MCP server from Comet ML for managing, visualizing, and tracking machine learning experiments directly within development environments.
What business challenges will you tackle with these powerful new capabilities?
About the authors
Mark Roy is a Principal Machine Learning Architect for AWS, helping customers design and build generative AI solutions. His focus since early 2023 has been leading solution architecture efforts for the launch of Amazon Bedrock, the flagship generative AI offering from AWS for builders. Mark’s work covers a wide range of use cases, with a primary interest in generative AI, agents, and scaling ML across the enterprise. He has helped companies in insurance, financial services, media and entertainment, healthcare, utilities, and manufacturing. Prior to joining AWS, Mark was an architect, developer, and technology leader for over 25 years, including 19 years in financial services. Mark holds six AWS certifications, including the ML Specialty Certification.
Eashan Kaushik is a Specialist Solutions Architect AI/ML at Amazon Web Services. He is driven by creating cutting-edge generative AI solutions while prioritizing a customer-centric approach to his work. Before this role, he obtained an MS in Computer Science from NYU Tandon School of Engineering. Outside of work, he enjoys sports, lifting, and running marathons.
Madhur Prashant is an AI and ML Solutions Architect at Amazon Web Services. He is passionate about the intersection of human thinking and generative AI. His interests lie in generative AI, specifically building solutions that are helpful and harmless, and most of all optimal for customers. Outside of work, he loves doing yoga, hiking, spending time with his twin, and playing the guitar.
Amit Arora is an AI and ML Specialist Architect at Amazon Web Services, helping enterprise customers use cloud-based machine learning services to rapidly scale their innovations. He is also an adjunct lecturer in the MS data science and analytics program at Georgetown University in Washington, D.C.
Andy Palmer is a Director of Technology for AWS Strategic Accounts. His teams provide Specialist Solutions Architecture skills across a number of speciality domain areas, including AIML, generative AI, data and analytics, security, network, and open source software. Andy and his team have been at the forefront of guiding our most advanced customers through their generative AI journeys and helping to find ways to apply these new tools to both existing problem spaces and net new innovations and product experiences.
Generate compliant content with Amazon Bedrock and ConstitutionalChain
Generative AI has emerged as a powerful tool for content creation, offering several key benefits that can significantly enhance the efficiency and effectiveness of content production processes such as creating marketing materials, image generation, content moderation etc. Constitutional AI and LangGraph‘s reflection mechanisms represent two complementary approaches to ensuring AI systems behave ethically – with Anthropic embedding principles during training while LangGraph implements them during inference/runtime through reflection and self-correction mechanisms. By using LanGraph’s Constitutional AI, content creators can streamline their workflow while maintaining high standards of user-defined compliance and ethical integrity. This method not only reduces the need for extensive human oversight but also enhances the transparency and accountability of content generation process by AI.
In this post, we explore practical strategies for using Constitutional AI to produce compliant content efficiently and effectively using Amazon Bedrock and LangGraph to build ConstitutionalChain for rapid content creation in highly regulated industries like finance and healthcare. Although AI offers significant productivity benefits, maintaining compliance with strict regulations are crucial. Manual validation of AI-generated content for regulatory adherence can be time-consuming and challenging. We also provide an overview of how Insagic, a Publicis Groupe company, integrated this concept into their existing healthcare marketing workflow using Amazon Bedrock. Insagic is a next-generation insights and advisory business that combines data, design, and dialogues to deliver actionable insights and transformational intelligence for healthcare marketers. It uses expertise from data scientists, behavior scientists, and strategists to drive better outcomes in the healthcare industry.
Understanding Constitutional AI
Constitutional AI is designed to align large language models (LLMs) with human values and ethical considerations. It works by integrating a set of predefined rules, principles, and constraints into the LLM’s core architecture and training process. This approach makes sure that the LLM operates within specified ethical and legal parameters, much like how a constitution governs a nation’s laws and actions.
The key benefits of Constitutional AI for content creation include:
- Ethical alignment – Content generated using Constitutional AI is inherently aligned with predefined ethical standards
- Legal compliance – The LLM is designed to operate within legal frameworks, reducing the risk of producing non-compliant content
- Transparency – The principles guiding the LLM’s decision-making process are clearly defined and can be inspected
- Reduced human oversight – By embedding ethical guidelines into the LLM, the need for extensive human review is significantly reduced
Let’s explore how you can harness the power of Constitutional AI to generate compliant content for your organization.
Solution overview
For this solution, we use Amazon Bedrock Knowledge Bases to store a repository of healthcare documents. We employ a Retrieval Augmented Generation (RAG) approach, first retrieving relevant context and synthesizing an answer based on the retrieved context, to generate articles based on the repository. We then use the open source orchestration framework LangGraph and ConstitutionalChain to generate, critique, and review prompts in an Amazon SageMaker notebook and develop an agentic workflow to generate compliance content. The following diagram illustrates this architecture.
This implementation demonstrates a sophisticated agentic workflow that not only generates responses based on a knowledge base but also employs a reflection technique to examine its outputs through ethical principles, allowing it to refine and improve its outputs. We upload a sample set of mental health documents to Amazon Bedrock Knowledge Bases and use those documents to write an article on mental health using a RAG-based approach. Later, we define a constitutional principle with a custom Diversity, Equity, and Inclusion (DEI) principle, specifying how to critique and revise responses for inclusivity.
Prerequisites
To deploy the solution, you need the following prerequisites:
- An AWS account
- Appropriate AWS Identity and Access Management (IAM) permissions to access an Amazon Simple Storage Service (Amazon S3) bucket, create Amazon Bedrock knowledge bases, and create a SageMaker notebook instance
Create an Amazon Bedrock knowledge base
To demonstrate this capability, we download a mental health article from the following GitHub repo and store it in Amazon S3. We then use Amazon Bedrock Knowledge Bases to index the articles. By default, Amazon Bedrock uses Amazon OpenSearch Serverless as a vector database. For full instructions to create an Amazon Bedrock knowledge base with Amazon S3 as the data source, see Create a knowledge base in Amazon Bedrock Knowledge Bases.
-
- On the Amazon Bedrock console, create a new knowledge base.
- Provide a name for your knowledge base and create a new IAM service role.
- Choose Amazon S3 as the data source and provide the S3 bucket storing the mental health article.
- Choose Amazon Titan Text Embeddings v2 as the embeddings model and OpenSearch Serverless as the vector store.
- Choose Create Knowledge Base.
Import statements and set up an Amazon Bedrock client
Follow the instructions provided in the README file in the GitHub repo. Clone the GitHub repo to make a local copy. We recommend running this code in a SageMaker JupyterLab environment. The following code imports the necessary libraries, including Boto3 for AWS services, LangChain components, and Streamlit. It sets up an Amazon Bedrock client and configures Anthropic’s Claude 3 Haiku model with specific parameters.
Define Constitutional AI components
Next, we define a Critique class to structure the output of the critique process. Then we create prompt templates for critique and revision. Lastly, we set up chains using LangChain for generating responses, critiques, and revisions.
Define a State class and refer to the Amazon Bedrock Knowledge Bases retriever
We define a LangGraph State class to manage the conversation state, including the query, principles, responses, and critiques:
Next, we set up an Amazon Bedrock Knowledge Bases retriever to extract the relevant information. We refer to the Amazon Bedrock knowledge base we created earlier to create an article based on mental health documents. Make sure to update the knowledge base ID in the following code with the knowledge base you created in previous steps:
Create LangGraph nodes and a LangGraph graph along with constitutional principles
The next section of code integrates graph-based workflow orchestration, ethical principles, and a user-friendly interface to create a sophisticated Constitutional AI model. The following diagram illustrates the workflow.
It uses a StateGraph to manage the flow between RAG and critique/revision nodes, incorporating a custom DEI principle to guide the LLM’s responses. The system is presented through a Streamlit application, which provides an interactive chat interface where users can input queries and view the LLM’s initial responses, critiques, and revised answers. The application also features a sidebar displaying a graph visualization of the workflow and a description of the applied ethical principle. This comprehensive approach makes sure that the LLM’s outputs are not only knowledge-based but also ethically aligned by using customizable constitutional principles that guide a reflection flow (critique and revise), all while maintaining a user-friendly experience with features like chat history management and a clear chat option.
Streamlit application
The Streamlit application component of this code creates an interactive and user-friendly interface for the Constitutional AI model. It sets up a side pane that displays a visualization of the LLM’s workflow graph and provides a description of the DEI principle being applied. The main interface features a chat section where users can input their queries and view the LLM’s responses.
The application maintains a chat history, displaying both user inputs and LLM responses, including the initial response, any critiques generated, and the final revised response. Each step of the LLM’s process is clearly labeled and presented to the user. The interface also includes a Clear Screen button to reset the chat history. When processing a query, the application shows a loading spinner and displays the runtime, providing transparency into the LLM’s operation. This comprehensive UI design allows users to interact with the LLM while observing how constitutional principles are applied to refine the LLM’s outputs.
Test the solution using the Streamlit UI
In the Streamlit application, when a user inputs a query, the application initiates the process by creating and compiling the graph defined earlier. It then streams the execution of this graph, which includes the RAG and critique/revise steps. During this process, the application displays real-time updates for each node’s execution, showing the user what’s happening behind the scenes. The system measures the total runtime, providing transparency about the processing duration. When it’s complete, the application presents the results in a structured manner within the chat interface. It displays the initial LLM-generated response, followed by any critiques made based on the constitutional principles, and finally shows the revised response that incorporates these ethical considerations. This step-by-step presentation allows users to see how the LLM’s response evolves through the constitutional AI process, from initial generation to ethical refinement. As mentioned, in the GitHub README file, in order to run the Streamlit application, use the following code:
For details on using a Jupyter proxy to access the Streamlit application, refer to Build Streamlit apps in Amazon SageMaker Studio.
Modify the Studio URL, replacing lab? with proxy/8501/.
How Insagic uses Constitutional AI to generate compliant content
Insagic uses real-world medical data to help brands understand people as patients and patients as people, enabling them to deliver actionable insights in the healthcare marketing space. Although generating deep insights in the health space can yield profound dividends, it must be done with consideration for compliance and the personal nature of health data. By defining federal guidelines as constitutional principles, Insagic makes sure that the content delivered by generative AI complies with federal guidelines for healthcare marketing.
Clean up
When you have finished experimenting with this solution, clean up your resources to prevent AWS charges from being incurred:
- Empty the S3 buckets.
- Delete the SageMaker notebook instance.
- Delete the Amazon Bedrock knowledge base.
Conclusion
This post demonstrated how to implement a sophisticated generative AI solution using Amazon Bedrock and LangGraph to generate compliant content. You can also integrate this workflow to generate responses based on a knowledge base and apply ethical principles to critique and revise its outputs, all within an interactive web interface. Insagic is looking at more ways to incorporate this into existing workflows by defining custom principles to achieve compliance goals.
You can expand this concept further by incorporating Amazon Bedrock Guardrails. Amazon Bedrock Guardrails and LangGraph Constitutional AI can create a comprehensive safety system by operating at different levels. Amazon Bedrock provides API-level content filtering and safety boundaries, and LangGraph implements constitutional principles in reasoning workflows. Together, they enable multi-layered protection through I/O filtering, topic restrictions, ethical constraints, and logical validation steps in AI applications.
Try out the solution for your own use case, and leave your feedback in the comments.
About the authors
Sriharsh Adari is a Senior Solutions Architect at Amazon Web Services (AWS), where he helps customers work backwards from business outcomes to develop innovative solutions on AWS. Over the years, he has helped multiple customers on data platform transformations across industry verticals. His core area of expertise include Technology Strategy, Data Analytics, and Data Science. In his spare time, he enjoys playing sports, binge-watching TV shows, and playing Tabla.
David Min is a Senior Partner Sales Solutions Architect at Amazon Web Services (AWS) specializing in Generative AI, where he helps customers transform their businesses through innovative AI solutions. Throughout his career, David has helped numerous organizations across industries bridge the gap between cutting-edge AI technology and practical business applications, focusing on executive engagement and successful solution adoption.
Stephen Garth is a Data Scientist at Insagic, where he develops advanced machine learning solutions, including LLM-powered automation tools and deep clustering models for actionable, consumer insights. With a strong background spanning software engineering, healthcare data science, and computational research, he is passionate to bring his expertise in AI-driven analytics and large-scale data processing to drive solutions.
Chris Cocking specializes in scalable enterprise application design using multiple programming languages. With a nearly 20 years of experience, he excels in LAMP and IIS environments, SEO strategies, and most recently designing agentic systems. Outside of work, Chris is an avid bassist and music lover, which helps fuel his creativity and problem-solving skills.
Minimize generative AI hallucinations with Amazon Bedrock Automated Reasoning checks
Foundation models (FMs) and generative AI are transforming enterprise operations across industries. McKinsey & Company’s recent research estimates generative AI could contribute up to $4.4 trillion annually to the global economy through enhanced operational efficiency, productivity growth of 0.1% to 0.6% annually, improved customer experience through personalized interactions, and accelerated digital transformation.
Today, organizations struggle with AI hallucination when moving generative AI applications from experimental to production environments. Model hallucination, where AI systems generate plausible but incorrect information, remains a primary concern. The 2024 Gartner CIO Generative AI Survey highlights three major risks: reasoning errors from hallucinations (59% of respondents), misinformation from bad actors (48%), and privacy concerns (44%).
To improve factual accuracy of large language model (LLM) responses, AWS announced Amazon Bedrock Automated Reasoning checks (in gated preview) at AWS re:Invent 2024. Through logic-based algorithms and mathematical validation, Automated Reasoning checks validate LLM outputs against domain knowledge encoded in the Automated Reasoning policy to help prevent factual inaccuracies. Automated reasoning checks is part of Amazon Bedrock Guardrails, a comprehensive framework that also provides content filtering, personally identifiable information (PII) redaction, and enhanced security measures. Together, these capabilities enable organizations to implement reliable generative AI safeguards—with Automated Reasoning checks addressing factual accuracy while other Amazon Bedrock Guardrails features help protect against harmful content and safeguard sensitive information.
In this post, we discuss how to help prevent generative AI hallucinations using Amazon Bedrock Automated Reasoning checks.
Automated Reasoning overview
Automated Reasoning is a specialized branch of computer science that uses mathematical proof techniques and formal logical deduction to verify compliance with rules and requirements with absolute certainty under given assumptions. As organizations face increasing needs to verify complex rules and requirements with mathematical certainty, automated reasoning techniques offer powerful capabilities. For example, AWS customers have direct access to automated reasoning-based features such as IAM Access Analyzer, S3 Block Public Access, or VPC Reachability Analyzer.
Unlike probabilistic approaches prevalent in machine learning, Automated Reasoning relies on formal mathematical logic to provide definitive guarantees about what can and can’t be proven. This approach mirrors the rigors of auditors verifying financial statements or compliance officers validating regulatory requirements, but with mathematical precision. By using rigorous logical frameworks and theorem-proving methodologies, Automated Reasoning can conclusively determine whether statements are true or false under given assumptions. This makes it exceptionally valuable for applications that demand high assurance and need to deliver unambiguous conclusions to their users.
The following workflow illustrates solver-based formal verification, showing both the process flow and algorithm for verifying formal system properties through logical analysis and SAT/SMT solvers.
One of the widely used Automated Reasoning techniques is SAT/SMT solving, which involves encoding a representation of rules and requirements into logical formulas. A logical formula is a mathematical expression that uses variables and logical operators to represent conditions and relationships. After the rules and requirements are encoded into these formulas, specialized tools known as solvers are applied to compute solutions that satisfy these constraints. These solvers determine whether the formulas can be satisfied—whether there exist values for variables that make the formulas true.
This process starts with two main inputs: a formal representation of the system (like code or a policy) expressed as logical formulas, and a property to analyze (such as whether certain conditions are possible or requirements can be met). The solver can return one of three possible outcomes:
- Satisfiable – The solver finds an assignment of values that makes the formulas true, proving that the system can satisfy the given requirements. The solver provides this assignment, which can serve as a concrete example of correct behavior.
- Unsatisfiable – The solver proves that no assignment exists that make all formulas true, proving that the requirements can’t be met. This often comes with information about which constraints are in conflict, helping identify the incorrect assumptions in the system.
- Unknown – In some cases, the solver might not be able to determine satisfiability within reasonable computational limits, or the encoding might not contain enough information to reach a conclusion.
This technique makes sure that you either get confirmation that the specific property holds (with a concrete example), proof that it can’t be satisfied (with information on conflicting constraints), or an indication that the problem needs to be reformulated or analyzed differently.
Key features of Automated Reasoning checks
Automated Reasoning checks offer the following key features:
- Mathematical validation framework – The feature verifies LLM outputs using mathematical logical deduction. Unlike probabilistic methods, it uses sound mathematical approaches to provide definitive guarantees about system behaviors within defined parameters.
- Policy-based knowledge representation – Organizations can create Automated Reasoning policies that encode their rules, procedures, and guidelines into structured, mathematical formats. Organizations can upload documents like PDFs containing HR guidelines or operational workflows, which are then automatically converted into formal logic structures. Policy changes are automatically versioned with unique Amazon Resource Names (ARNs), allowing for change tracking, auditing, and rollback capabilities to maintain consistent policy enforcement.
- Domain expert enablement – The feature is designed to empower domain experts, such as HR personnel or operational managers, to directly encode their knowledge without technical intermediaries. This makes sure that business rules and policies are accurately captured and maintained by those who understand them best.
- Natural language to logic translation – The system uses two complementary approaches: LLMs handle natural language understanding, and a symbolic reasoning engine performs mathematical validation. This hybrid architecture allows users to input policies in plain language while maintaining mathematically rigorous verification.
- Explainable validation results – Each validation check produces detailed findings that indicate whether content is Valid, Invalid, or No Data. The feature provides clear explanations for its decisions, including extracted factual statements, and suggested corrections for invalid content.
- Interactive testing environment – Users can access a chat playground on the Amazon Bedrock console to test and refine policies in real time. The feature supports both interactive testing through the Amazon Bedrock console and automated testing through API integrations, with the ability to export test cases in JSON format for integration into continuous testing pipelines or documentation workflows.
- Seamless AWS integration – The feature integrates directly with Amazon Bedrock Guardrails and can be used alongside other configurable guardrails like Contextual Grounding checks. It can be accessed through both the Amazon Bedrock console and APIs, making it flexible for various implementation needs.
These features combine to create a powerful framework that helps organizations maintain factual accuracy in their AI applications while providing transparent and mathematically sound validation processes.
Solution overview
Now that we understand the key features of Automated Reasoning checks, let’s examine how this capability works within Amazon Bedrock Guardrails. The following section provides a comprehensive overview of the architecture and demonstrates how different components work together to promote factual accuracy and help prevent hallucinations in generative AI applications.
Automated Reasoning checks in Amazon Bedrock Guardrails provides an end-to-end solution for validating AI model outputs using mathematically sound principles. This automated process uses formal logic and mathematical proofs to verify responses against established policies, offering definitive validation results that can significantly improve the reliability of your AI applications.
The following solution architecture follows a systematic workflow that enables rigorous validation of model outputs.
The workflow consists of the following steps:
- Source documents (such as HR guidelines or operational procedures) are uploaded to the system. These documents, along with optional intent descriptions, are processed to create structured rules and variables that form the foundation of an Automated Reasoning policy.
- Subject matter experts review and inspect the created policy to verify accurate representation of business rules. Each validated policy is versioned and assigned a unique ARN for tracking and governance purposes.
- The validated Automated Reasoning policy is associated with Amazon Bedrock Guardrails, where specific policy versions can be selected for implementation. This integration enables automated validation of generative AI outputs.
- When the generative AI application produces a response, Amazon Bedrock Guardrails triggers the Automated Reasoning checks. The system creates logical representations of both the input question and the application’s response, evaluating them against the established policy rules.
- The Automated Reasoning check provides detailed validation results, including whether statements are Valid, Invalid, or No Data. For each finding, it explains which rules and variables were considered, and provides suggestions for making invalid statements valid.
With this solution architecture in place, organizations can confidently deploy generative AI applications knowing that responses will be automatically validated against your established policies using mathematically sound principles.
Prerequisites
To use Automated Reasoning checks in Amazon Bedrock, make sure you have met the following prerequisites:
- An active AWS account
- Access permission through your AWS Account Manager, because Automated Reasoning checks is currently in gated preview
- Confirmation of AWS Regions where Automated Reasoning checks is available
Input dataset
For this post, we examine a sample Paid Leave of Absence (LoAP) policy document as our example dataset. This policy document contains detailed guidelines covering employee eligibility criteria, duration limits, application procedures, and benefits coverage for paid leave. It’s an ideal example to demonstrate how Automated Reasoning checks can validate AI-generated responses against structured business policies, because it contains clear rules and conditions that can be converted into logical statements. The document’s mix of quantitative requirements (such as minimum tenure and leave duration) and qualitative conditions (like performance status and approval processes) makes it particularly suitable for showcasing the capabilities of automated reasoning validation.
The following screenshot shows an example of our policy document.
Start an Automated Reasoning check using the Amazon Bedrock console
The first step is to encode your knowledge—in this case, the sample LoAP policy—into an Automated Reasoning policy. Complete the following steps to initiate an Automated Reasoning check using the Amazon Bedrock console:
- On the Amazon Bedrock console, choose Automated Reasoning Preview under Safeguards in the navigation pane.
- Choose Create policy.
- Provide a policy name and policy description.
- Upload your source document. The source content can’t be modified after creation and must not exceed 6,000 characters with limitations on table sizes and image processing.
- Include a description of the intent of the Automated Reasoning policy you’re creating. For the sample policy, you can use the following intent:
The policy creation process takes a few minutes to complete. The rules and variables are created after creating the policy and they can be edited, removed, or have new rules or variables added to them.
The policy document version is outlined in the details section along with the intent description and build status.
Next, you create a guardrail in Amazon Bedrock by configuring as many filters as you need.
- On the Amazon Bedrock console, choose Guardrails under Safeguards in the navigation pane.
- Choose Create guardrail.
- Provide guardrail details such as a name and an optional description.
- Add an Automated Reasoning check by choosing Enable Automated Reasoning policy, and choose the policy name and version.
- Choose Next and complete the creation of the guardrail.
- Navigate back to the Automated Reasoning section of the Amazon Bedrock console and open the newly created policy. You can use the test playground and input sample questions and answers that represent real user interactions with your LLM.
- Choose the guardrail you created, then choose Submit to evaluate how your policy handles these exchanges.
After submitting, you’ll be presented with one or more findings. A finding contains a set of facts that were extracted from the input Q&A and are analyzed independently. Each finding includes four key components:
- Validation results – Shows the outcome of Automated Reasoning checks. The system determines these results by evaluating extracted variable assignments against your defined policy rules.
- Applied rules – Displays the specific rules from your policy that were used to reach the validation conclusion.
- Extracted variables – Lists the variables that were identified and used in the validation process.
- Suggestions – Shows variable assignments that would make invalid responses valid, or for valid responses, identifies necessary assumptions. These can be used to generate feedback for your LLM.
Finally, you can use the feedback suggestions to improve your LLM’s responses.
- Collect rules from valid results with suggestions and invalid results.
- Feed these collected variables and rules back to your LLM to revise its original.
- Refine your policy:
- Edit incorrect rules using natural language.
- Improve variable descriptions when Automated Reasoning checks fail to assign values.
- For effective variable descriptions, include both technical definitions and common user expressions. For example, for a variable named
is_full_time, "works more than 20 hours per week"is technically correct because it’s a quote from the source policy, but won’t help Automated Reasoning checks understand what users mean when they say “part-time.” Instead, use"works more than 20 hours per week; set to true if user says 'full-time' and false if user says 'part-time'".
Start an Automated Reasoning check using Python SDK and APIs
First, you need to create an Automated Reasoning policy from your documents using the Amazon Bedrock console as outlined in the previous section. Next, you can use the policy created with the ApplyGuardrail API to validate your generative AI application.
To use the Python SDK for validation using Automated Reasoning checks, follow these steps:
- First, set up the required configurations:
- Before using Amazon Bedrock with Automated Reasoning policies, you will need to load the required service models. After being allowlisted for Amazon Bedrock access, you will receive two model files along with their corresponding version information. The following is a Python script to help you load these service models:
- After you set up the service models, initialize the AWS clients for both Amazon Bedrock and Amazon Bedrock Runtime services. These clients will be used to interact with the models and apply guardrails.
- Before applying Automated Reasoning policies, you need to either locate an existing guardrail or create a new one. The following code first attempts to find a guardrail by name, and if not found, creates a new guardrail with the specified Automated Reasoning policy configuration. This makes sure you have a valid guardrail to work with before proceeding with policy enforcement.
- When testing guardrails with Automated Reasoning policies, you need to properly format your input data. The following code shows how to structure a sample question and answer pair for validation:
- Now that you have your formatted input data, you can apply the guardrail with Automated Reasoning policies to validate the content. The following code sends the input to Amazon Bedrock Guardrails and returns the validation results:
- After applying guardrails, you need to extract and analyze the Automated Reasoning assessment results. The following code shows how to process the guardrail output:
The output will look something like the following:
When a response violates AR policies, the system identifies which rules were violated and provides information about the conflicts. The feedback from the AR policy validation can be routed back to improve the model’s output, promoting compliance while maintaining response quality.
Possible use cases
Automated Reasoning checks can be applied across various industries to promote accuracy, compliance, and reliability in AI-generated responses while maintaining industry-specific standards and regulations. Although we have tested these checks across multiple applications, we continue to explore additional potential use cases. The following table provides some applications across different sectors.
| Industry | Use Cases |
| Healthcare |
|
| Financial Services |
|
| Travel and Hospitality |
|
| Insurance |
|
| Energy and Utilities |
|
| Manufacturing |
|
Best practices for implementation
Successfully implementing Automated Reasoning checks requires careful attention to detail and a systematic approach to achieve optimal validation accuracy and reliable results. The following are some key best practices:
- Document preparation – Use structured text-based PDF documents. Content should be limited to 6,000 characters. Avoid complex formatting that could interfere with the logical model generation.
- Intent description engineering – Create precise policy intents using a clear format. The intent should comprehensively cover expected use cases and potential edge cases. For example:
- Policy validation – Review the generated rules and variables to make sure they accurately capture your business logic and policy requirements. Regular audits of these rules help maintain alignment with current business policies.
- Comprehensive testing –Develop a diverse set of sample Q&As in the test playground to evaluate different validation scenarios (valid, valid with suggestions, and invalid responses). Include edge cases and complex scenarios to provide robust validation coverage.
- Iterative improvement –Regularly update rules and LLM applications based on validation feedback, paying special attention to suggested variables and invalid results to enhance response accuracy. Maintain a feedback loop for continuous refinement.
- Version control management – Implement a systematic approach to policy versioning, maintaining detailed documentation of changes and conducting proper testing before deploying new versions. This helps track policy evolution and facilitates rollbacks if needed.
- Error handling strategy – Develop a comprehensive plan for handling different validation results, including specific procedures for managing invalid responses and incorporating suggested improvements into the response generation process.
- Runtime optimization – Understand and monitor the two-step validation process (fact extraction and logic validation) to achieve optimal performance. Regularly review validation results to identify patterns that might indicate needed improvements in variable descriptions or rule definitions.
- Feedback integration – Establish a systematic process for collecting and analyzing validation feedback, particularly focusing on cases where NO_DATA is returned or when factual claims are incorrectly extracted. Use this information to continuously refine variable descriptions and policy rules.
Conclusion
Amazon Bedrock Automated Reasoning checks represent a significant advancement in formally verifying the outputs of generative AI applications. By combining rigorous mathematical validation with a user-friendly interface, this feature addresses one of the most critical challenges in AI deployment: maintaining factual consistency and minimizing hallucinations. The solution’s ability to validate AI-generated responses against established policies using formal logic provides organizations with a powerful framework for building trustworthy AI applications that can be confidently deployed in production environments.
The versatility of Automated Reasoning checks, demonstrated through various industry use cases and implementation approaches, makes it a valuable tool for organizations across sectors. Whether implemented through the Amazon Bedrock console or programmatically using APIs, the feature’s comprehensive validation capabilities, detailed feedback mechanisms, and integration with existing AWS services enable organizations to establish quality control processes that scale with their needs. The best practices outlined in this post provide a foundation for organizations to maximize the benefits of this technology while maintaining high standards of accuracy.
As enterprises continue to expand their use of generative AI, the importance of automated validation mechanisms becomes increasingly critical. We encourage organizations to explore Amazon Bedrock Automated Reasoning checks and use its capabilities to build more reliable and accurate AI applications. To help you get started, we’ve provided detailed implementation guidance, practical examples, and a Jupyter notebook with code snippets in our GitHub repository that demonstrate how to effectively integrate this feature into your generative AI development workflow. Through systematic validation and continuous refinement, organizations can make sure that their AI applications deliver consistent, accurate, and trustworthy results.
About the Authors
Adewale Akinfaderin is a Sr. Data Scientist–Generative AI, Amazon Bedrock, where he contributes to cutting edge innovations in foundational models and generative AI applications at AWS. His expertise is in reproducible and end-to-end AI/ML methods, practical implementations, and helping global customers formulate and develop scalable solutions to interdisciplinary problems. He has two graduate degrees in physics and a doctorate in engineering.
Nafi Diallo is a Sr. Applied Scientist in the Automated Reasoning Group and holds a PhD in Computer Science. She is passionate about using automated reasoning to ensure the security of computer systems, improve builder productivity, and enable the development of trustworthy and responsible AI workloads. She worked for more than 5 years in the AWS Application Security organization, helping build scalable API security testing solutions and shifting security assessment left.
AWS App Studio introduces a prebuilt solutions catalog and cross-instance Import and Export
AWS App Studio is a generative AI-powered service that uses natural language to build business applications, empowering a new set of builders to create applications in minutes. With App Studio, technical professionals such as IT project managers, data engineers, enterprise architects, and solution architects can quickly develop applications tailored to their organization’s needs—without requiring deep software development skills. Common use cases range from inventory management and approval workflows to content management and operational portals, and beyond – App Studio adapts to streamline a wide variety of business applications.
Since the general availability of App Studio in November 2024, customers across diverse industries have adopted it to build scalable, enterprise-grade applications, transforming their development processes and accelerating time-to-market. App Studio customers, including both enterprises and system integrators, have shared the need for portability and reusability across App Studio instances. Based on their experience, two areas of interests emerged:
- Getting started – New customers and builders asked to learn and explore the product through readily available examples and patterns that explain application building possibilities in App Studio.
- Optimizing time to value – Teams often validate use cases in a sandbox before moving to production. This highlights an interest in a more efficient approach to share and deploy applications across multiple App Studio instances.
Today, App Studio announced two new features to accelerate application building:
- Prebuilt solutions catalog – Featuring a set of practical examples and common patterns (like S3 and Bedrock integration) to accelerate getting started and enable deployment of applications from the catalog to production environments in less than 15 minutes.
- Cross-instance Import and Export – Enabling straightforward and self-service migration of App Studio applications across AWS Regions and AWS accounts.
In this post, we walk through how to use the prebuilt solutions catalog to get started quickly and use the Import and Export feature
Prerequisites
To follow along with this post, you should have the following prerequisites:
- Access to App Studio. For more information, see Setting up and signing in to App Studio.
- Optional: Review App Studio concepts to familiarize yourself with important App Studio concepts.
- Optional: An understanding of basic web development concepts, such as JavaScript syntax.
- Optional: Familiarity with AWS services.
Prebuilt solutions catalog
App Studio is introducing a prebuilt solutions catalog to accelerate the way builders approach application building. This resource offers a diverse collection of prebuilt applications that can be seamlessly imported into your App Studio instance, serving as both a learning tool and a rapid deployment solution. By providing access to proven patterns and prebuilt solutions, App Studio significantly reduces the initial setup time for builders, enabling you to move from concept to production in less than 15 minutes.
The catalog includes a variety of practical use cases including a Product Adoption Tracker to manage customer feedback, track feature requests, and summarize meeting notes with AI. To import the Product Adoption Tracker, navigate to the prebuilt solutions catalog, copy an import code, and follow the import instructions in the next section.
Import an application
You now have the ability to import an App Studio application from a different App Studio instance. Importing applications is available to all builders and admins.
Complete the following steps to import an App Studio application:
- Sign in and launch the App Studio instance where you want to import an application.
- Choose My applications in the navigation pane.
- Choose the dropdown menu next to Create app and choose Import app.
- Enter an import code from the prebuilt app catalog or that you generated by the export process outlined in the next section and choose Import. Depending on the application size, you might need to wait a few seconds for the import to finish.
- After completion, the application will be imported to your development environment. You can explore the debug panel at the bottom of the page to understand which custom connectors need to be connected to automations and entities.
Now that we have successfully imported an application, let’s walk through how we can export our own applications to a different App Studio instance.
Export an application
You now have the ability to export an App Studio application to a different App Studio instance. Generating an application export creates a static snapshot with all artifacts needed to recreate the application—automations, components, and entities. After importing, you will need to reconnect custom connectors to automations and entities.
Application security and control are maintained through a robust permissions system. Only authorized application owners and co-owners can generate application exports and restrict which App Studio instances can import a given application. If needed, application owners can revoke access by deactivating the import link at any time.
To export an App Studio application, complete the following steps:
-
- Sign in to the App Studio instance that you want to export an application from.
- Choose My applications in the navigation pane.
- Choose the dropdown menu next to Edit and choose Export.
- To restrict which App Studio instances can import this application, configure application import permissions:
- Anyone with the import code can import this application – Grant import permissions to all instances. Only select this option if you want anyone with the import code to have access to import your application.
- Only specified App Studio instances can import this application – Provide the specific instance IDs that can import the application (multiple instances can be separated by commas). To find your instance ID, navigate to your instance’s account settings by choosing Account settings on the App Studio console.
- Choose Generate import code to generate a unique import code.
- Two additional options for managing import codes are available after the application has been exported at least once to application owners and co-owners:
- Generate new import code – When you make updates to this application, you will need to generate a new import code by choosing Generate new code. Generating a new code invalidates the old code, but will not automatically refresh existing imported applications.
- Delete import code – To stop application import access, choose this option. Deleting the import code will invalidate the current code and prevent subsequent import attempts. Applications previously created using this code will continue to work.
Considerations
The following are some key considerations for using the prebuilt solutions catalog and importing and exporting applications across App Studio instances:
- There is no cost associated with importing and exporting applications, including importing applications from the prebuilt solutions catalog.
- Applications cannot be imported into the same instance, but you can achieve a similar result of replicating functionality within an instance by duplicating apps, components, and pages.
- There are no limits on the number of applications you can import or export. The maximum number of applications in an App Studio instance is subject to service quotas.
Conclusion
Jumpstart your app building workflow with App Studio’s prebuilt solutions catalog and Import and Export features. Effortlessly migrate applications across AWS instances, collaborate with teams, and transfer applications to clients. Start using App Studio’s prebuilt solutions catalog and Import and Export features today – we’re excited to see how you will use these features to accelerate your application building journey.
To learn more about App Studio, explore more features on the App Studio page. Get started with App Studio in the AWS Management Console. Experience the App Studio workshop for hands-on learning, and join the conversation in the #aws-app-studio channel in the AWS Developers Slack workspace.
Read more about App Studio
- Build and modify apps using natural language with App Studio, now generally available
- Discover how customers are innovating with App Studio
Watch App Studio demos
- The Fastest and Easiest Way to Build Business Applications
- Watch how the NFL accelerated image processing with App Studio
About the Authors
Umesh Kalaspurkar is a Principal Solutions Architect at AWS based in New York, bringing over two decades of expertise in digital transformation and innovation across both enterprise and startup environments. He specializes in designing solutions that help organizations overcome their most pressing challenges. When not architecting cloud solutions, Umesh cherishes time spent with his children, carving down ski slopes, and exploring new destinations around the world.
Samit Kumbhani is an AWS Senior Solutions Architect in the New York City area with over 18 years of experience. He currently partners with independent software vendors (ISVs) to build highly scalable, innovative, and secure cloud solutions. Outside of work, Samit enjoys playing cricket, traveling, and biking.
Haoran (Hao) Su is a Senior Technical Account Manager in New York City with over 8 years of experience with the cloud. He collaborates with Software, Internet and Model providers (SWIM) and Digitally Native Businesses (DNB) to improve their financial and operational efficiency, and architectural resiliency. Outside of work, Hao enjoys international traveling, exercising, and streaming.
Anshika Tandon is a Senior Product Manager – Technical at AWS with a decade of experience building AI and B2B SaaS products from concept to launch. She excels in cross-functional product leadership, focusing on delivering measurable business value through strategic initiatives. A global citizen having lived in 10 cities and visited 26 countries, Anshika balances her professional life with interests in skiing, travel, and performing in improv comedy shows.
Alex (Tao) Jia is a Senior Product Marketing Manager at AWS, focusing on generative AI. With 15+ years in tech marketing, she drives products from concept to scale, shaping positioning, fostering adoption, and leading global go-to-market strategies. She has worked with enterprises and ISVs, reaching millions of developers. Outside work, Alex enjoys exploring technology’s impact on humanity through books, research, and conversations.