Kyc aml · Production

Indonesian digital signature provider automates KYC identity verification on-premises with Nanonets AI

The problem

The client needed to verify citizen IDs for every new user to meet KYC compliance standards, but processing photos of physical IDs with traditional OCR was unreliable—taking approximately 3 minutes per document manually—causing slow turnaround and customer loss. Privacy requirements also mandated an on-premises deployment.

First attempt

Traditional OCR tools could not separate the ID from junk data in the background of uploaded photos, resulting in low accuracy and requiring manual reprocessing of documents.

Workflow diagram · grounded in source
1
User uploads physical ID photo
trigger
“users had to upload a photo of their physical ID for the KYC requirements”
2
AI identifies and extracts ID data
ai_action
“They could train an AI model to identify the ID card accurately and then pull the required data”
3
On-premises compliance integration
integration
“In partnership with PT Sai, we implemented an on-premises solution that complied with the data protection standards”
4
Onboarding process completed
output
“This enabled them to simplify their onboarding process and launch the product”
Reported outcome

Nanonets' on-premises AI solution enabled compliant automated identity verification, simplified the onboarding process, and allowed the client to launch and offer their product.

Reported metrics
Manual document processing time (baseline)~3 minutes per document manually
Customer retention impact from slow verificationlosing customers
Identity verification automationautomating identity verification effectively
Reported stack
NanonetsOCRPT Sai
Source
https://nanonets.com/customer-success-story/indonesian-digital-signature-provider-automates-kyc-process-on-premises
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Nanonets' on-premises AI solution enabled compliant automated identity verification, simplified the onboarding process, and allowed the client to launch and offer their product.

What tools did this team use?

Nanonets, OCR, PT Sai.

What results were reported?

Manual document processing time (baseline): ~3 minutes per document manually; Customer retention impact from slow verification: losing customers; Identity verification automation: automating identity verification effectively (source-reported, not independently verified).

What failed first in this deployment?

Traditional OCR tools could not separate the ID from junk data in the background of uploaded photos, resulting in low accuracy and requiring manual reprocessing of documents.

How is this kyc aml AI workflow structured?

User uploads physical ID photo → AI identifies and extracts ID data → On-premises compliance integration → Onboarding process completed.