(US & Canada) Mark Birkhead, Firmwide CDO at JPMorganChase, speaks with Sajid Khan, Partner/Principal, AI and Data Financial Services at EY, in a video interview about the modernization strategy for technology and platforms, the expectations from generative AI investments, and the approach to selecting AI use cases.
Birkhead explains that the bank’s modernization strategy has been a multi-year process that hinges on strong partnerships with external providers who play a vital role. Given the wide range of ideas, startups, and established players in this space, the bank prioritizes caution, driven by stringent standards around cybersecurity, data storage, access controls, and entitlements. It recognizes the need for precision, even if it means taking time to complete the modernization journey.
A central aspect of the strategy is adopting a multi-cloud infrastructure to avoid over-reliance on any single provider, thus reducing risk if technology shifts. This multi-cloud approach, combined with on-premises solutions, is seen as essential for future readiness. Birkhead notes that platforms are being built also to accommodate an AI-driven world, where intelligent agents could autonomously navigate applications in the same way humans do. Preparing for this future requires extensive reconfiguration, governance, and controls, which the team is actively developing.
Further, Birkhead reveals that the bank is also heavily focused on building platforms capable of real-time data streaming. With a strong emphasis on real-time data availability, it aims to serve clients who need instant updates. This capability is especially crucial for areas like Chase businesses, payments, fraud detection, and risk management, where real-time information enables the bank to deliver up-to-the-millisecond insights across all client interactions.
When asked about expectations from GenAI investments, Birkhead explains that his team has been deeply invested in natural language processing (NLP) for years, especially for high-volume businesses like Chase, which handles millions of customer calls every month. With the introduction of large language models (LLMs) and smaller language models (SLMs), he sees these technologies as a natural evolution of the NLP efforts. By using these advanced models, the organization aims to improve both operational efficiency and customer satisfaction, even though the impacts may not always be easily quantifiable in dollar terms.
For example, in Chase’s call centers, LLMs and NLP models support agents by instantly retrieving relevant documents and information based on real-time customer interactions. This capability saves valuable seconds on each call, ultimately benefiting both the company and its customers by reducing wait times and improving service quality. Birkhead points out that when these time-saving benefits scale across the entire firm, the cumulative effect on efficiency and customer experience is substantial.
Moreover, Birkhead highlights that the team is using these models to better understand customer satisfaction levels, enabling insights into customer sentiment and providing feedback on their services. Last year, the organization launched an internal AI tool called "LLM Suite," which has now been deployed to 150,000 employees. This tool assists with tasks like document summarization, writing, problem-solving, and idea generation. Importantly, the platform operates within a secure environment, equipped with guardrails to protect sensitive information, respect privacy, and comply with data governance standards.
Looking forward, Birkhead believes these more focused models, which are rising in popularity, can provide even more targeted and efficient solutions, allowing businesses of all sizes to leverage AI in a customized, effective way.
Additionally, Birkhead believes the alignment of various interests creates an opportunity to leverage AI capabilities to positively impact customers while ensuring safety and promoting beneficial outcomes. He emphasizes that AI can support better governance and responsible practices, aligning with both regulatory expectations and the company’s values.
However, he notes the importance of focusing on the right use cases, stressing JPMorgan Chase's deliberate approach to selecting and developing AI initiatives. Not every idea is pursued, as they invest significant effort in evaluating each use case to ensure it meets risk and appropriateness standards for the organization.
CDO Magazine appreciates Mark Birkhead for sharing his insights with our global community.