How One Bank Built Its Own AI In-House

A Vietnamese bank is building its own AI tools with open-source models to keep sensitive customer data secure and avoid big spending.

VietBank is taking a different approach to artificial intelligence by building its capabilities from the ground up. Under CIO NghiaTran, the bank has implemented a lean AI plan that favors in-house development with open-source tools over expensive contracts with large technology vendors. The primary driver for this strategy is data security; as a financial institution, VietBank cannot allow sensitive customer information to leave its own secure environment. This has led to the creation of several internal AI-powered systems, including a smart office tracking system and tools for advanced customer intelligence. By integrating AI with its CRM and behavioral analytics, the bank aims to build a strategic advantage without compromising on its strict data governance policies. This self-reliant model gives them complete control over their technology stack and allows them to tailor solutions directly to their unique business needs.

This strategy serves as a valuable case study for CTOs, IT leaders, and security teams, particularly those operating in highly regulated sectors like finance and healthcare. VietBank’s success demonstrates that impactful AI innovation is possible without massive budgets or reliance on third-party cloud platforms for core model processing. By self-hosting large language models (LLMs) and other open-source tools, organizations can directly address critical concerns around data sovereignty, privacy, and regulatory compliance. This in-house approach not only mitigates security risks but also enables greater agility. Teams can rapidly develop, deploy, and iterate on custom AI applications that solve specific internal challenges, offering a powerful, cost-effective, and secure alternative to off-the-shelf solutions.

This is a valuable case study for CTOs and IT leaders on implementing a lean, in-house AI strategy using open-source tools, particularly in a regulated industry. It highlights key considerations like data security, rapid deployment, and the benefits of self-hosting LLMs.

VietBank's approach shows how companies in regulated industries can leverage AI without relying on expensive third-party vendors, giving them greater control over data security, customization, and costs. This model allows for faster development of tailored solutions, potentially creating a competitive advantage through enhanced customer intelligence and operational efficiency.