What challenges did Sun Finance face before AI implementation?

Manual identity verification was time-consuming, labor-intensive, and error-prone, causing delays and inconsistencies in customer onboarding. Customers often lacked high bandwidth internet or digital services, complicating access and verification on low-end devices. Sun Finance needed a scalable, automated solution to speed ID verification, reduce fraud, and improve customer experience globally.

What was the impact of the AI solution?

Customer identity verification processing accelerated to 15-20 seconds, enabling near real-time onboarding. Manual effort and errors in identity verification dramatically reduced. Up to 60% of applications are automatically approved in certain markets, expanding financial inclusion and improving risk management.

Source

ProductionEvidence: Medium65/100

Sun Finance Uses Amazon Rekognition to Combat Identity Fraud and Remove Customer Friction

Use case typeFraud detectionUpdated Jun 13, 2026

Sun Finance, a Latvian fintech company, used to rely on a manual, time-consuming, error-prone identity verification (IDV) process for customer onboarding. The challenge was improving speed, accuracy, and fraud prevention while expanding financial services access globally, including to underserved regions with limited internet and device capabilities. In 2019, Sun Finance automated their IDV workflow by implementing Amazon Rekognition to compare customer selfies with ID documents and detect potential fraud or duplicate accounts. They also adopted Amazon Textract to accurately extract text from documents, including vertically or angled written content. The solution processes identity verification in near real-time, typically completing in 15-20 seconds, significantly reducing manual effort. This automation improved customer onboarding speed and allowed automatic application approvals up to 60% in some markets, enhancing financial inclusion and risk mitigation.

Organization: Sun Finance
Industry: Finance
Location: Latvia
Published: April 2026

Reported outcomes

15-20 seconds

timeTime & speed

60%quantified impact

Strategic outcomes

Speed & agilityNear-real-time identity verificationRisk & complianceImproved fraud detection and risk managementCustomer experience & trustReduced onboarding frictionScale & capacityExpanded automatic application approvals

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

Normalized claim

Time: 15-20 seconds

AWS Customer StoriesApr 29, 2026Customer storyInferred claimMedium evidence strength

Customer identity verification processing accelerated to 15-20 seconds, enabling near real-time onboarding.

Normalized claim

Quantified impact: 60%

AWS Customer StoriesApr 29, 2026Customer storyInferred claimMedium evidence strength

Up to 60% of applications are automatically approved in certain markets, expanding financial inclusion and improving risk management.

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

Customer: Sun Finance
Provider: AWS
Maturity: Production
Linked source: AWS Customer Stories

Deployed Amazon Rekognition for facial recognition to verify selfies against official ID documents, automating identity checks

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 3 of 3

1Identity Verification
2Fraud Detection
3Customer Onboarding

Manual identity verification was time-consuming, labor-intensive, and error-prone, causing delays and inconsistencies in customer onboarding.
Customers often lacked high bandwidth internet or digital services, complicating access and verification on low-end devices.
Sun Finance needed a scalable, automated solution to speed ID verification, reduce fraud, and improve customer experience globally.

Deployed Amazon Rekognition for facial recognition to verify selfies against official ID documents, automating identity checks.
Used Amazon Textract for accurate document text extraction, including difficult cases such as vertical or angled text.
Implemented multiple Amazon Rekognition features to detect fraudulent accounts, ensure selfie capture quality, and maintain a private face vector database for similarity matching.

Technologies

Amazon Rekognition Amazon Textract

Customer identity verification processing accelerated to 15-20 seconds, enabling near real-time onboarding.
Manual effort and errors in identity verification dramatically reduced.
Up to 60% of applications are automatically approved in certain markets, expanding financial inclusion and improving risk management.

Sources & evidence1

Evidence: Medium65/100Evidence strength