The Future of Insurance Podcast – Guru Vasudeva
SVP & CTO of P&C, Nationwide
Season 9, Episode 9, October 13, 2026
Episode Details
Guru Vasudeva is Senior Vice President and Chief Technology Officer for Property & Casualty at Nationwide, where he leads technology strategy for one of the largest P&C operations in the U.S. He joined Nationwide in 2004 and has held a progression of senior technology roles there, including Chief Architect, Business Unit CIO, Enterprise Chief Technology Officer, and SVP & CIO for Program & Application Services, before moving into his current P&C-wide CTO role.
Before Nationwide, Guru was an Executive Architect at IBM and a software engineer at AT&T. He holds a Bachelor of Engineering in Computer Science from the University of Mysore (1992) and an MBA in International Finance from the University of Cincinnati College of Business (2000).
He is currently leading Nationwide’s $1.5 billion, multi-year technology and AI investment across the P&C business — spanning underwriting, claims, software engineering, and customer experience.
Highlights from the Show
Three Layers of AI at Nationwide:
- Everyday AI: A Nationwide-specific generative AI tool rolled out in full at the start of the year to 25,000-plus associates and another 5,000 to 10,000 contractors, including the ability to build their own daily agents.
- AI for tech: Started in software engineering and is expanding into operations and cybersecurity.
- Business transformation: Underwriting, Claims, and customer service, which Guru calls the hardest layer because it requires picking something bold.
- Context: Nationwide has run near real-time predictive analytics at scale for 15 years, so generative AI is the new capability, not AI itself.
Software Engineering at Scale:
- Train the trainer: Across roughly 400 development teams, three people each (tech leads, Scrum Masters, and app owners) were trained on autonomous development and MCP gateways, then taught their teams.
- Results: 10 to 15 percent efficiency in aggregate, with some use cases near 40 percent.
- Human change: Guru calls the people side the harder part, since gains disappear when one teammate uses the tools well and another does not.
- Scale: About $500 million of Nationwide’s annual technology spend goes to software engineering projects, with close to 5,000 associates and contractors.
Jevons Paradox in Software:
- Demand grows: As software gets cheaper, a backlog of business requests becomes viable. Guru committed to return $10 to 15 million to P&C leadership, and most of it is being reinvested in new projects.
- Legacy migration: A project estimated at $5 million under the old methodology is now about $750,000 and on track to finish before year end.
- The mother of all spreadsheets: One engineer and one underwriter turned a sprawling middle market rules spreadsheet into an online tool in a couple of days.
- COBOL to Angular: An 80,000-line COBOL codebase was migrated in three days, and the engineer was recognized at Nationwide’s annual technology conference.
Underwriting as a Growth Opportunity:
- The submission problem: Complex submissions run hundreds of pages in inconsistent formats, information is missing midway through, and renewals consume much of an underwriter’s time.
- What changed: Earlier ingestion had to be hard-coded. Generative AI can read documents from different sources, recognize that different words say the same thing, and synthesize them.
- The goal: Judge whether a file is decision ready, help collect what’s missing, triage by appetite, and let underwriters manage double the territory.
- Ambition: Guru wants Nationwide known as the most responsive P&C carrier in the Industry.
Bilingual Talent:
- The gap: Too few technologists know the business at detail level, and too few business people know AI.
- Guru’s view: It is easier to teach business people AI than to teach AI people the business, especially now that knowledge is abundant and agents can speed up the work.
- Niche partners: Nationwide is finding strong unlocks with smaller companies that sit at the intersection of insurance and AI.
Guardrails for Everyday AI:
- Access controls: Rolling out to IT leadership first exposed a long-standing issue: the tool made it far easier to discover files people could technically access, so permissions had to be tightened.
- Private tenant: Nationwide’s data stays in its own tenant, and the vendor agreement bars use of that information.
- Agents and shadow IT: As associates build agents, governance is the open question, and traditional methods come first.
- The genie is out of the box: Without a protected internal tool, people will use personal accounts anyway.
Looking Ahead:
- Prototypes in minutes: The constraint is now imagination, and Guru says projects are finishing on schedule and under budget more often than at any point in his 30 years in IT.
- The new constraint: Faster coding does not help if business and technology cannot agree on requirements, so work structures need to evolve.
- Next breakthroughs: Underwriting triage, surge call handling after a CAT event, and fraud detection and lifecycle automation in Claims.
- Startup community: Guru spends significant time with VC firms and startups to find solutions that can jumpstart business transformation.
Watch the Episode
Listen to the Episode
Thank you to our sponsor
This episode is brought to you by The Future of Auto Insurance: Connected, Embedded, Subscribed, which you can download for free today, thanks to support from Guidewire.
This report is part of The Future of Insurance book series, available globally from Amazon in print, Kindle and Audible audiobook.
Like the music in the show?
Stream or download music from Hyperbeat Music on your favorite platform.


