What was the impact of the AI solution?

Reduced visa processing time to under 4 minutes. Automated around 95% of short-term visa applications, supporting up to 1,000 per day. Improved operational efficiency by up to 99 percent and strengthened border security.

Source

Scaled productionEvidence: Medium65/100

Papua New Guinea Immigration and Citizenship Services Authority cuts visa processing to under 4 minutes with generative AI on AWS

Use case typeIntelligent document processingUpdated Jun 13, 2026

Papua New Guinea Immigration and Citizenship Services Authority (ICSA) partnered with AWS partner NiuPay to modernize manual, error-prone visa processing. The solution analyzes scanned documents and images with generative AI and structured data extraction, integrates directly into the eVisa platform, and uses configurable rules for auto-approval or human escalation.

Organization: Papua New Guinea Immigration and Citizenship Services Authority
Industry: Public Sector
Location: Papua New Guinea
Published: May 2026

Reported outcomes

4 minutes

timeTime & speed

95%quantified impact+99%productivity

Strategic outcomes

Speed & agilityAccelerated visa processing workflowScale & capacityAutomated short-term visa applications at scaleRisk & complianceStrengthened border securityNew product / capabilityIntegrated AI into eVisa platform

Why do we believe this?Outcome claims, sources, and evidence checks

Normalized claim

Time: 4 minutes decrease

AWS Customer StoriesMay 27, 2026Customer storyInferred claimMedium evidence strength

Reduced visa processing time to under 4 minutes.

Normalized claim

Quantified impact: 95%

AWS Customer StoriesMay 27, 2026Customer storyInferred claimMedium evidence strength

Automated around 95% of short-term visa applications, supporting up to 1,000 per day.

Normalized claim

Productivity: 99% increase

AWS Customer StoriesMay 27, 2026Customer storyInferred claimMedium evidence strength

Improved operational efficiency by up to 99 percent and strengthened border security.

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: Papua New Guinea Immigration and Citizenship Services Authority
Provider: AWS
Maturity: Scaled Production
Linked source: AWS Customer Stories

ICSA needed to extract accurate information from diverse text and image document formats at scale

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 3 of 3

1Document Processing
2Workflow Automation
3Intelligent Automation

Manual visa processing took days to weeks and had data entry validation gaps.
ICSA needed to extract accurate information from diverse text and image document formats at scale.
The agency wanted to improve border security while accelerating tourism and business travel.

Built MyICSA Analysis Engine on AWS with Amazon Bedrock and Amazon Textract.
Integrated the system directly into the eVisa platform.
Used an orchestrator/decision engine with configurable rules that non-technical staff can update.

Technologies

Amazon Bedrock Amazon Textract

Reduced visa processing time to under 4 minutes.
Automated around 95% of short-term visa applications, supporting up to 1,000 per day.
Improved operational efficiency by up to 99 percent and strengthened border security.

Architecture

MyICSA Analysis Engine is a generative AI document processing system built on AWS and integrated into the eVisa platform. It uses Amazon Bedrock for NLP, Amazon Textract for structured extraction from scanned documents, and a CV-based photo validation component, with a configurable rules-based orchestrator that auto-approves confident cases and escalates uncertain ones to human review.

Implementation partners1

NiuPay

Sources & evidence1

Evidence: Medium65/100Evidence strength