TD Bank AI Agent Cuts Mortgage Decision Time by 15 Hours

Title

TD Bank AI Agent Cuts Mortgage Decision Time by 15 Hours

Content

TD Bank Group's AI team built and deployed its first AI agent in January for residential mortgage loan decisions, completing application review in minutes rather than the 15 hours typically required by a human reviewer. The agent was deployed internally across TD's mortgage loan officer team, which has embraced the tool. Chad Koziel, Assistant Vice President at TD and founder of the AI startup Layer 6 (acquired by TD in 2018), explained the operational shift: 'If you just imagine people who spent a lot of time doing some fairly arduous tasks, looking through various IDs, hunting through 100-page documents for that key piece of information, we can reshape their work into something that's much more closely aligned with delivering on the brand promise that TD has of more human.' TD chose mortgage processing as the first agentic AI use case because it is a sharply defined task with clear inputs (purchase agreement, government-issued ID, account statements, proof of income), clear outputs, and a human in the loop. 'For our first use case, we wanted to make sure it was something that made a massive positive impact for our clients, so the residential lending journey fit the bill,' Koziel told American Banker. Technically, the AI agent handles document ingestion, verification, and inconsistency checks, then presents a summary to a credit adjudicator for final decision. The agent does not replace rules-based logic alone because, Koziel explained, 'It is not possible to reliably account for all of these differences between humans using rules-based models, that's why we had humans do the process.' Generative AI is capable of this nuanced reasoning. TD declined to name the single foundation model underpinning the agent but uses a combination of proprietary models, open-source models, and off-the-shelf foundation models including Anthropic's Claude and OpenAI's ChatGPT. At operational scale, TD underwrites hundreds of thousands of mortgages every year, with $826 million in total U.S. residential mortgage origination volume in 2025. External experts Patrick Hall (chief AI officer at George Washington University School of Business) and Brad Leimer (founder and principal at Leimer One Advisors) noted the deployment's cross-functional design — combining subject-matter experts from the business, scientists from Layer 6, engineers, and risk and control partners — as a model for embedding AI into core operating workflows. Future direction: TD is exploring additional agentic AI use cases following this first transformation, with the mortgage deployment serving as a reference pattern for further workflow redesign.