Scaling Quality with AI: How We Built Agent-Based QA and a Secure Internal GPT

As financial systems grow in scale and regulatory complexity, traditional QA approaches struggle to keep pace with the volume of requirements, risks, and test artifacts that must be continuously reviewed and maintained. In regulated fintech environments, QA teams must balance speed, accuracy, and compliance—often relying on manual effort that does not scale. This session presents a real-world case study of how the Acba Bank QA organization evolved from manual, human-heavy processes to an AI-assisted quality ecosystem built around purpose-driven AI agents and a secure, in-house GPT platform. The talk explores how they designed and introduced AI agents for requirements review, risk identification, test plan creation, test case design, automation script generation, and intelligent summary reporting; all grounded in internal knowledge sources such as Confluence, Jira, and TestRail. A key focus of the session will be on why and how they implemented an internal GPT solution instead of relying on public AI tools, enabling secure interaction with sensitive documentation, test artifacts, and automation assets while meeting enterprise security and governance requirements. Attendees will leave with practical guidance on architecture choices, governance models, and adoption strategies, along with clear lessons on how AI can augment—not replace—QA expertise to improve quality outcomes at scale.