Data Requisition Copilot: 60% Faster Data Access for a Leading Bank 

The client is a global bank having millions of customers across retail, corporate, and banking areas. With data stored in various legacy and new systems, the bank aimed to give its teams quicker access to data while meeting strict regulatory and security requirements. 

The Challenge: Slow & IT-Dependent Data Access

• Legacy applications owned different parts of the tolling lifecycle and shared limited interoperability
• Data sat in multiple systems, making end-to-end visibility difficult
• Manual checks and reconciliations slowed down processing • Rigid system architectures made it hard to scale with rising traffic volumes
• User access and governance were inconsistent across systems

The Solution: Easy Data Access with DRCopilot

Datamatics built an Agentic AI-powered Data Requisition Copilot that allows users to request data through simple natural language yet ensures all data stays within the bank’s secure infrastructure.

Key Design Principle: Security First

Although users interact via Microsoft Teams, no data is accessed, stored, or passed through Teams. All data extraction, validation, processing, and delivery happen within the bank’s private environment, with Teams acting only as a conversational interface. This ensured zero external data exposure, full compliance with banking regulations and end-to-end auditability & governance. The solution was delivered in two phases.

Phase 1: Foundation Setup

• Requirement workshops with cross-functional teams
• Implemented prioritized data requests on the Enterprise Data Warehouse
• Enabled NLQ-to-SQL conversion for secure, direct database queries
• Established governance rules, approval workflows, and audit logs

Phase 2: Enterprise Scale-Up

• Extended capability to multiple source systems
• Introduced automated validation, exception handling, and dataset preparation
• Connected backend systems via APIs and secure database layers, all within the bank’s controlled environment

• Unified API Platform: A centralized API layer was introduced to manage the entire toll process, from transaction generation to reconciliation. This replaced several independent legacy modules
• Role-Based Access Control: A structured access model provided consistent governance and security across user groups. This addressed issues left by older systems
• Modern, Responsive User Interface: A new, more intuitive user interface with real-time insights and simplified daily workflows replaced the outdated screens
• Cloud-Native Deployment on GCP: The platform operated on Google Cloud Platform, using Kubernetes offering high availability, scalability, and operational resilience that the old infrastructure couldn’t provide
• DevOps-enabled CI/CD Pipeline: A CI/CD pipeline to simplified deployments, cut down on manual work, and added predictability to release cycles
• Integration with Enterprise Systems: The new platform enabled connection between AS400 and the Enterprise Data Warehouse ensuring smooth data exchange without harming ongoing operations