We partnered with Viston to accelerate digital onboarding, strengthen Anti-Money Laundering (AML) processes, and provide an AI-powered copilot for frontline agents while meeting local data residency requirements.

Business Challenge

Customer onboarding averaged 5 days, manual KYC processes created significant operational workload, fraud attempts continued to rise, and frontline agents frequently provided inconsistent policy guidance.

Our Approach and Solution

Viston developed a document intelligence pipeline featuring OCR, entity extraction, and automated cross-checks, along with liveness detection, face matching, and a graph-based AML risk scoring system.

Digital Onboarding Automation

Implemented AI-powered document processing with identity verification, OCR, entity extraction, and automated validation to streamline customer onboarding.

Graph-Based AML Monitoring

Built a graph-based AML risk scoring solution to improve fraud detection and identify suspicious financial relationships.

AI Agent Copilot

Deployed an AI-powered Agent Copilot using Retrieval-Augmented Generation (RAG) over internal policies and product documentation, delivering grounded and auditable responses directly within the CRM.

AI Applications and Benefits Delivered

Intelligent Document Processing

AI-powered ID document parsing, OCR, and automated verification to accelerate customer onboarding.

AML & Fraud Detection

Sanctions screening, Politically Exposed Person (PEP) and Beneficial Ownership (PBO) verification, along with transaction anomaly detection.

AI Knowledge Assistant

Knowledge-grounded conversational AI for frontline agents with policy citations and contextual recommendations.

Business Benefits

• Faster customer onboarding
• Reduced false positives
• Shorter customer service calls
• Improved compliance traceability
• Enhanced agent productivity

Cost of Implementation

USD 135,000, including data residency controls, MLOps implementation, and AI model validation.

Time to Implement

20 weeks to full production deployment across retail banking onboarding and the contact center.

Tools and Technologies Used

• Azure OpenAI (UAE North)
• Azure Form Recognizer
• Azure Face API
• Neo4j
• Databricks
• Elasticsearch
• Kubernetes
• Kong API Gateway
• HashiCorp Vault
• Microsoft Power BI

Quantitative Outcomes

• 64% reduction in onboarding time (5 days → 1.8 days)
• 29% reduction in fraud losses
• 17% improvement in AML detection precision
• 22% reduction in agent Average Handle Time (AHT)
• 13-point improvement in First Contact Resolution (FCR)
• Zero critical audit findings across two compliance reviews

Key Performance Indicators (KPIs) Tracked

• Onboarding turnaround time (TAT)
• KYC/AML alert precision and recall
• Suspicious Activity Report (SAR) volume
• False positive rate
• Agent Average Handle Time (AHT)
• First Contact Resolution (FCR)
• Policy grounding citation rate

Pre- and Post-Implementation Metrics

• Average Handle Time (AHT): 7.9 minutes → 6.2 minutes
• False Positive Rate: 42% → 29%
• Grounding Citation Presence: 0% → 100% for Copilot responses
• SAR Conversion to Investigation: +21%

Stakeholder Quotes

Head of Retail Banking

“The copilot’s citations changed the game with auditors. Every recommendation is traceable.”

Regulatory and Compliance Considerations

• UAE data residency requirements
• ISO 27001 compliance
• PCI DSS network segmentation
• FATF guidance alignment
• Explainable AI
• Model retention policies