CDO Magazine Podcast Series · CDO Magazine

PODCAST | Truist AI Leader on Why Efficiency Is the First Real Win for Enterprise AI

April 9, 2026·11 min·1 clip

Sanjay Sankoli reveals why AI wins are in workflow acceleration, not autonomous decisions, across front, middle, and back offices.

Karan Jain, founder and CEO of Naya One, opens by interviewing Sanjay Sankoli, Chief Architect for AI and Data at Truist Bank. Sankoli frames the current AI success story as one of augmentation and workflow acceleration rather than autonomous decision-making. Across the front office, he points to customer service deflection, predictive servicing, retention, and underwriting augmentation as concrete win areas. In the middle office, fraud detection refinements, KYC and AML dispositions, and claims processing are benefiting from generative AI and agentic workflows. Back office gains come from document intelligence that extracts structured information from unstructured sources, enabling downstream decision augmentation. Developer productivity is singled out as a horizontal AI enabler with significant and measurable enterprise impact. Sankoli argues the common thread across all wins is decision augmentation paired with workflow acceleration. Jain agrees and adds his own observation that these gains are appearing across clients at scale. The conversation turns to the split between bottom-line efficiency and top-line revenue as AI motivators. Sankoli states that regulatory pressure and organizational confidence constraints currently restrict AI's top-line contribution. Jain notes that 18 months ago top-line AI discussions were absent, but are now emerging, particularly in wealth management and digital channels for cross-selling. He speculates that meaningful top-line impact may not materialize until after 2026. Sankoli then addresses what data architecture constraints most limit enterprise AI at scale. He describes how decades of M&A activity and regulatory patchwork have created siloed, fragmented data infrastructure across capital markets, payments, and banking. These 'automated islands' trap organizational context and intelligence, making it difficult to harness coherent enterprise-level understanding. Data fragmentation is compounded by legacy core systems that treat data as a project asset rather than an enterprise asset. Real-time processing remains immature compared to batch-optimized setups, and cloud strategies have not yet reached the maturity needed to support compute-intensive AI. Vendor lock-in and platform sprawl add further constraints. Sankoli emphasizes that rationalization — simplifying, standardizing, and making data AI-ready — is a prerequisite for meaningful AI scale. Jain closes by identifying the push-pull between rationalization and AI investment as a rising tension that C-suite leaders across organizations are actively navigating.

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from CDO Magazine Podcast Series