GXBank

Caroline Chong

Data Leadership Authority

GXBank was built cloud-native from day one, leveraging real-time data flows as our core growth engine rather than retrofitting digital channels onto legacy branch networks.

In this environment, effective data leadership bridges technical infrastructure with commercial outcomes across three vital levers: ecosystem-led inclusion, which has enabled us to scale to over 1.6 million customers. As of Q1 2026, we disbursed approx. RM53 million in MSME credit via GX Biz FlexiLoan, proactive risk operations that evaluate non-traditional data streams, and the ability to decouple transaction growth from headcount through intelligent automation.

Balancing AI Innovation with Human Judgment

We intentionally position AI as an operational co-pilot, not as a substitute for human judgment in regulated decision-making. Tools we developed in-house are designed to lower operating costs and allow analysts and business teams to focus on higher-value work and innovate at speed.  A prime example is BI Bytes, our proprietary GenAI semantic chatbot that embeds verified banking logic to allow non-technical teams to query Snowflake SQL in plain English with 94% accuracy. By absorbing 30% of ad-hoc reporting requests, BI Bytes has saved an estimated RM800,000 over three years - while keeping all models grounded in a single source of truth within our central data warehouse.

"As leaders, our job is to spotlight the steady hands behind the scenes, from data governance to data privacy, which ensure our data is secure, accurate and compliant."

However, scale and speed are meaningless without trust. Processing over 450 million transactions as of Q1 of 2026 naturally increases the risk operations complexity. As fraudsters increasingly turn to AI and automation to scale their attacks, static rules fall short. To stay ahead, we counter by deploying "AI vs. AI" capabilities to intercept evolving threats in real time.

Turning AI into an Operational Advantage

We maintain responsible data practices by deliberately introducing "good friction," like contextual scam warnings when high-risk behavior is detected - slowing things down just enough to stop fraud without impacting user experience. We also enforce strict human accountability through our dual-engine AI framework (FrAIdy and TrAIdy), where AI synthesizes complex histories while human analysts make all final regulated decisions. For lending, our GuardPlus tool runs sub-0.1-second document metadata analysis to combat lending fraud instantly.

As digital banks mature, data teams must shift from initial setup to managing complex scale without headcount inflation. For instance, FrAIdy and TrAIdy automate multi-system data retrieval, saving 16,000 working hours annually and slashing case processing times from 15-20 minutes down to 1-3 minutes. This allows us to scale to more than 1.6 million customers with zero additional risk ops headcount. Navigating this shift requires leadership that relentlessly solves for true business value, automates manual workflows to free up team expertise, and builds AI as an empowering co-pilot that strengthens human judgment rather than replacing it.

Building High-Performing Data Teams

Ultimately, people drive innovation. Throughout our journey at GXBank, I have learned that high-performing data teams are not built on cutting-edge algorithms alone. They are built on a rock-solid foundation and psychological empowerment.

As leaders, our job is to spotlight the steady hands behind the scenes, from data governance to data privacy, which ensure our data is secure, accurate and compliant. When you grant your team the trust and creative freedom to solve complex problems, anchored by the purpose of fiercely protecting our customers, you build an environment where people do not just execute; they innovate with pride.