top of page
Amaris  Logo_Horizontal(HighRes3)-06.png


Automated detection and Redaction Software powered by AI

Automated Detection and Redaction

Use groundbreaking AI to automatically detect and redact sensitive data, at rest or in transit.

Cheap and Efficient

Cost effective and easy installation of software.


Only authorised users with tokens can access redacted data, keeping it secure at-rest and at-transit.

Automated Redaction Tool

investigation (1).png


Select &

Select files/folders to process. Choose types of PII Entities to redact. 



Detect &

Tool automatically detects and anonymizes PII. 

pencil (1).png


Review & 
Correct Results

Review, add or change redactions if required. 

download (3).png


Sava Data & 

Generate Report

Tool saves data in chosen format e.g. PDF, JSON. Generates anonymization report if required.

Personally Identifiable Information (PII)

AI Detection

Uses state of the art AI to detect anything that can be deemed as PII.

Pattern Recognition

AI detects repeating patterns within documents highlighting PII, for faster recognition of similar information for quick redaction

Manual Finetuning

Human in the loop system to enable manual correction of AI Redaction, to improve the system overall 

Optical Character Recognition

 AI detects images that can be deemed as PII and redacts them accordingly

Personally Identifiable Information (PII) 



Replace personally identifiable information with characters to obscure original text. 

Original NRIC


Masked NRIC





Replace different PII with unique substitute values.

  • Same token represents each instance of the original data. 

  • Option for token reference to be stored and encrypted is available. 

AI models can ingest token as substitute words without reading any PII directly which reduces bias and prevents data leakage into the model weight. 

  • Original values can be covered by Authorized Users to obtain model insights. 

Review & Correct Results

Batch Checking

  • Collate list of detected PII from batch of documents.

  • Highlight lower confidence detections for user to verify.  

Actively Learned Corrections

  • User identifies misassigned entities and reassigns them 

  • Reassignments are used to train the PII detection model and improve detections over time. 

Handling highly sensitive data

check-mark (1).png

Containment of data within the Intranet

check-mark (1).png

De-identification and encryption of data-at-rest and data-in-transit

check-mark (1).png

Controlled access to only authorised personnel

check-mark (1).png

Complete anonymisation of data and generation of synthetic data for use in unsecured instances

Use Cases

Screenshot 2022-01-13 114922.png


Anonymizer helps to protect PII of patients, sensitive information in documents during file transfers, digitally and physically.

Screenshot 2022-01-13 115144.png


Anonymizer helps to protect key PII of clients/customers and of organisations involved in banking activities.

bottom of page