DBS Bank - Group Institutional Banking Operations (IBGO): Centralised AI Platform Qorus-Infosys Finacle Banking Innovation Awards 2025- Nominated

IBGO built a centralised AI platform to address manual transactions and associated workloads, which were inefficient. Using Generative AI (GenAI), Large Language Models (LLMs), and Optical Character Recognition (OCR++), we have automated everything from trade finance document processing and cash operations forms to corporate loan data entry and customer service interactions. The initial focus was on internal applications to build expertise before moving towards customer-facing solutions. A cross-functional governance structure and multi-disciplinary teams were established to drive the project. Customer focus groups and stakeholder engagement were also undertaken, alongside establishing a secure technological foundation for LLM integration

Concept and Objectives: The project aims to eliminate the tail of manual transactions across key operational areas covering: Trade Financing, Cash Operations, Corporate Loans/Servicing, and Customer Servicing. The objective is to enhance operational efficiency, reduce errors, and improve customer experience by leveraging AI and automation.

Reasons Behind the Project: Despite a decade of successfully driving digital transformation and automation efforts, a substantial volume of manual transactions persisted deemed as un-addressable, hindering further operational efficiency gains, the specific challenges of which included: • Trade Financing: Physical document handling and unstructured data led to inefficiencies and errors. • Cash Operations: Manual form processing caused errors and increased investigations, negatively impacting customer experience. • Corporate Loans/Servicing: Manual data entry and cross-system validation created inefficiencies and delays. • Servicing: Limited internet banking self-service functionality resulted in high call volumes and manual data entry, compounded by a chatbot with limited capabilities.

State of the Competition: Our competition faces similar challenges with manual transaction processing. DBS-Bank aims to differentiate itself by implementing a comprehensive AI-driven solution with robust control mechanisms and a focus on customer experience.

Sources of Inspiration: The project builds upon the existing "Managing through Journeys" (MtJ) customer-centric approach and integrates new technology experimentation, data-driven control towers provide real-time insights into business drivers and customer indicators, guiding AI development priorities and outcomes.

Departments Involved: The project involves a cross-functional collaboration between Operations, Strategy & Transformation, Technology, Business and Risk. A senior-level governance body oversees the project, while multi-disciplinary teams execute the initiatives. Customer focus groups and stakeholder engagement play a key role through-out the program.

Main Results So Far: Throughout 2024, the project focused on addressing initial opportunities in Corporate Loans, Servicing, and Cash Operations, with Trade Financing being tackled later.

The initial focus was on internal applications to build expertise before moving towards customer-facing solutions. A cross-functional governance structure and multi-disciplinary teams were established to drive the project. Customer focus groups and stakeholder engagement were also undertaken, alongside establishing a secure technological foundation for LLM integration.