FOI Redaction

Try Before You Buy

About NDX

NDX (National Digital Exchange) is a government platform that helps public sector organisations discover, evaluate, and adopt digital solutions.

Overview

Learning Artifact: This is a pre-deployed demonstration environment for learning and exploration, not a production-ready product. Use it to understand how AWS AI services can automate document redaction for compliance requirements.

Deployed and ready - no setup required. When you request this sandbox, you’ll receive access to a fully configured document redaction system that’s already running and ready to explore.

This FOI Redaction scenario demonstrates how local authorities can use AWS AI services to automatically identify and redact sensitive personal information from documents before release under Freedom of Information requests.

What you’ll explore

This sandbox includes a working implementation using:

Amazon Comprehend - Natural language processing to identify personal data entities (names, addresses, phone numbers and more)
Amazon Textract - Optical character recognition for extracting text from scanned documents and images
AWS Lambda - Serverless compute for orchestrating the redaction workflow
Amazon S3 - Storage for documents before and after redaction

In plain English: You’ll explore an AI system that automatically removes sensitive information from documents - like names, addresses, and phone numbers - so they can be safely released in response to FOI requests.

Learn More

Before requesting your sandbox, explore the scenario documentation to understand what’s deployed and how to interact with it:

View FOI Redaction scenario details

Getting started

Select “Try this now for 24 hours” above
Accept the Acceptable Use Policy
Receive AWS SSO credentials via email
Access your sandbox environment through AWS Console
Follow the scenario guide to upload test documents and see redaction in action

Why this matters for local government

Automated FOI redaction can:

Speed up response times by reducing manual document review
Improve consistency in identifying what needs to be redacted
Reduce compliance risk by catching sensitive data that humans might miss
Free up staff time for reviewing edge cases and complex requests

This scenario lets you evaluate whether AI-assisted redaction could work for your council’s FOI processes.

Constraints

Budget limit: $50 maximum spend (sufficient for extensive exploration)
Duration: 24 hours from activation
Purpose: Evaluation only (non-production use)
No cost to you - the budget is provided by NDX:Try

Important Notes

Resource Lifecycle Warning: All resources in your sandbox will be automatically deleted when your session time expires OR when the budget limit is reached (whichever comes first). Do not store important data in this environment.

Before your session ends

Export any configurations or code snippets you want to keep
Take screenshots of redaction results
Document any insights about accuracy and edge cases
Download any test outputs

How this was built

This scenario is built with open source infrastructure as code. You can view the CloudFormation template, understand how it works, and adapt it for your own needs.

View the source code on GitHub (opens in new tab)

Build your own

After exploring this pre-deployed scenario, use the Empty Sandbox to build your own solution from scratch. The Empty Sandbox gives you a clean AWS environment where you can experiment freely.

Explore more scenarios

Empty AWS Sandbox - Start fresh with a clean AWS environment
Council Chatbot - AI-powered resident services
Planning AI - Intelligent document analysis
Text to Speech - Accessibility audio generation

Troubleshooting

Try button not working? Make sure you’re signed in to NDX
Did not receive credentials? Check your spam folder, or wait a few minutes
Cannot access the scenario? Ensure you’re using the correct AWS region (us-east-1)
Budget exhausted early? Some services have higher costs - contact support if unexpected

Support

For technical issues during your sandbox session, contact the NDX team at ndx@dsit.gov.uk.