This sample is a document review solution powered by generative AI (Amazon Bedrock). It streamlines review processes involving extensive documents and complex checklists using a Human in the Loop approach. It supports the entire process from checklist structuring to AI-assisted review and final human judgment, reducing review time and improving quality.
Important
This tool is intended only for decision support and does not provide professional judgment or legal advice. All final judgments must be made by qualified human experts.
Warning
This sample may undergo breaking changes without prior notice.
Efficiently verify that product development specifications meet requirements and industry standards. Automate the process of comparing thousands of specifications annually against hundreds of checkpoints. AI extracts and structures relevant information from specifications, visualizing compliance results. Reviewers can efficiently perform final verification.
Verify that complex technical manuals comply with internal guidelines and industry standards. Support the process of comparing tens of thousands of pages of technical documentation annually against thousands of quality criteria. Automatically detect missing technical information and inconsistencies, supporting the creation of consistent, high-quality manuals.
Check that procurement documents and proposals meet necessary requirements. Automatically extract required information from documents spanning hundreds of pages, streamlining thousands of document reviews annually. Improve procurement process speed and accuracy by having humans verify compliance results against requirement lists.
There are two methods for deployment:
This method allows you to deploy directly from your browser using AWS CloudShell without preparing a local environment.
-
Enable Amazon Bedrock Models
Access Bedrock Model Access from the AWS Management Console and enable access to the following models:
- Anthropic Claude 3.7 Sonnet
- Amazon Nova Premier
By default, the Oregon (us-west-2) region is used, but you can change it with the
--bedrock-regionoption. -
Open AWS CloudShell
Open AWS CloudShell in the region where you want to deploy.
-
Run the Deployment Script
wget -O - https://raw.githubusercontent.com/aws-samples/review-and-assessment-powered-by-intelligent-documentation/main/bin.sh | bashThis one-liner command automatically executes everything from repository cloning to deployment.
-
Specify Custom Parameters (Optional)
wget -O - https://raw.githubusercontent.com/aws-samples/review-and-assessment-powered-by-intelligent-documentation/main/bin.sh | bash -s -- --ipv4-ranges '["192.168.0.0/16"]'
Available options:
--ipv4-ranges: IPv4 address ranges to allow in the frontend WAF (JSON array format)--ipv6-ranges: IPv6 address ranges to allow in the frontend WAF (JSON array format)--disable-ipv6: Disable IPv6 support--auto-migrate: Whether to automatically run database migration during deployment--cognito-self-signup: Whether to enable self-signup for the Cognito User Pool (true/false)--cognito-user-pool-id: Existing Cognito User Pool ID (creates new if not specified)--cognito-user-pool-client-id: Existing Cognito User Pool Client ID (creates new if not specified)--cognito-domain-prefix: Prefix for the Cognito domain (auto-generated if not specified)--mcp-admin: Whether to grant admin privileges to the MCP runtime Lambda function (true/false)--bedrock-region: Region to use for Amazon Bedrock (default: us-west-2)--document-model: AI model ID for document processing (default: us.anthropic.claude-3-7-sonnet-20250219-v1:0)--image-model: AI model ID for image review processing (default: us.anthropic.claude-3-7-sonnet-20250219-v1:0)--repo-url: URL of the repository to deploy--branch: Branch name to deploy--tag: Deploy a specific Git tag
-
Post-Deployment Verification
Upon completion of the deployment, the frontend URL and API URL will be displayed. Access the displayed URL to start using the application.
Important
With this deployment method, if you do not set option parameters, anyone who knows the URL can sign up. For production use, we strongly recommend adding IP address restrictions and disabling self-signup (--cognito-self-signup=false).
- Clone this repository
git clone https://github.com/aws-samples/review-and-assessment-powered-by-intelligent-documentation.git
- Prepare the backend
cd review-and-assessment-powered-by-intelligent-documentation
cd backend
npm ci
npm run prisma:generate
npm run build
- Install CDK packages
cd ../cdk
npm ci
- Edit parameter.ts as needed. See Parameter Customization for details.
- Before deploying CDK, you need to bootstrap once for the target region.
npx cdk bootstrap
- Deploy the sample project
npx cdk deploy --require-approval never --all
- You will see output like the following. Access the Web application URL displayed in
RapidStack.FrontendURLfrom your browser.
✅ RapidStack
✨ deployment time: 78.57s
Output:
...
RapidStack.FrontendURL = https://xxxxx.cloudfront.netThe following parameters can be customized during CDK deployment:
| Parameter Group | Parameter Name | Description | Default Value |
|---|---|---|---|
| WAF Configuration | allowedIpV4AddressRanges | IPv4 ranges to allow in the frontend WAF | ["0.0.0.0/1", "128.0.0.0/1"] (all allowed) |
| allowedIpV6AddressRanges | IPv6 ranges to allow in the frontend WAF | ["0000::/1", "8000::/1"] (all allowed) | |
| Cognito Settings | cognitoUserPoolId | Existing Cognito User Pool ID | Create new |
| cognitoUserPoolClientId | Existing Cognito User Pool Client ID | Create new | |
| cognitoDomainPrefix | Cognito domain prefix | Auto-generated | |
| cognitoSelfSignUpEnabled | Whether to enable self-signup for Cognito User Pool | true (enabled) | |
| Migration | autoMigrate | Whether to automatically run migration during deployment | true (auto-run) |
| MCP Features | mcpAdmin | Whether to grant admin privileges to the MCP runtime Lambda function (details) | false (disabled) |
| Citations API | enableCitations | Whether to enable Citations API for PDF documents (AWS announcement) | true (enabled) |
| Map State Concurrency | reviewMapConcurrency | Map State concurrency for the Review Processor (must be configured in consultation with throttling limits) | 1 |
| Map State Concurrency | checklistInlineMapConcurrency | Inline Map State concurrency for the Checklist Processor (must be configured in consultation with throttling limits) | 1 |
This application uses Strands agents with tools such as file reading, so you must select models that support tool use.
Examples of tool use supported models:
mistral.mistral-large-2407-v1:0(Mistral Large 2)us.anthropic.claude-3-5-sonnet-20241022-v2:0(Claude 3.5 Sonnet)us.amazon.nova-premier-v1:0(Amazon Nova Premier)
Important Notes:
-
Cross-region inference profiles: When using cross-region inference, regional prefixes like
us.,eu.,apac.are required for model IDs -
Official Documentation: Supported models and model features - Amazon Bedrock
Configuration Example:
// cdk/lib/parameter.ts
export const parameters = {
documentProcessingModelId: "mistral.mistral-large-2407-v1:0", // Mistral Large 2
bedrockRegion: "ap-northeast-1", // Tokyo region
// ...
};To configure these, directly edit the cdk/lib/parameter.ts file.
Caution
For production environments, it is strongly recommended to set cognitoSelfSignUpEnabled: false to disable self-signup. Leaving self-signup enabled allows anyone to register an account, which may pose a security risk.
By default, the autoMigrate parameter is set to true, which automatically runs database migrations during deployment. For production environments or environments containing important data, consider setting this parameter to false and controlling migrations manually.
- Developer Guide: Technical specifications, architecture, development environment setup
See CONTRIBUTING for more information.
This project is distributed under the license described in LICENSE.