The Data & AI Chief

Understand how to close the gap between AI experimentation and enterprise production. Shub Agarwal, Founder of the AI Trust Lab at USC and author of Successful AI Product Creation: A Nine-Step Framework, shares his AI product management framework for taking enterprise AI strategy from demo to production, drawing on two decades of product leadership at Amazon and Fortune 50 firms. He breaks down why experimentation must tie directly to business OKRs, the four mindset shifts leaders need to scale AI responsibly, and how the AI Trust Lab is building a benchmark evaluation framework for AI model trust and governance.

Key Moments:

Why 80% of AI Projects Never Reach Production (02:13): Shub traces the root cause of stalled AI programs to a missing system for moving from demo to deployment. Most teams have no repeatable path to production.

Shub's Nine-Step Framework for Building AI Products (06:00): Most AI projects start with a cool model instead of a painful problem. Shub walks through the three phases of his framework: discovery, execution, and excellence.

The Case Against "Fix Your Data First" (12:41): Conventional wisdom says clean your data before building AI. Shub challenges that, arguing modern LLMs offer far more flexibility with imperfect data.

Four Mindset Shifts for Scaling Enterprise AI (16:35): Shub outlines the four shifts separating organizations that scale AI from those that stall, from measuring AI performance differently to embedding trust from day one.

Inside Shub's AI Trust Lab at USC (23:54): Major foundation models are already being benchmarked on trust and safety. Shub explains the lab's mission to build a standardized evaluation framework for AI model governance.

Why Enterprise AI Governance Needs Multiple Disciplines (28:36): AI models can be sycophantic, manipulative, or lack candor. Shub argues that building trustworthy AI demands an interdisciplinary approach.

Key Quotes:

“I think the fundamental problem that organizations are facing today… is not that they have a lack of experimentation in the demo aspect. The challenge is they don't know how to take those demos to production, and that is where I saw the gap.” - Shub Agarwal

“I do think data is the fuel for AI… But I think today organizations are crippled by this ‘fix your data, and then we'll build AI’, and they never build AI. They never build use cases that are adding value.” - Shub Agarwal

“There's no FICO scores for models, so I decided to create one. I built this lab… bringing the computer scientists, the researchers, the applied AI researchers, the policy, and the communication people together to think of what is trust, define it, and ultimately measure and evaluate it.” - Shub Agarwal

Mentions

USC AI Trust Hub

Successful AI Product Creation: A Nine-Step Framework by Shub Agarwal

Four Steps to Epiphany: Successful Strategies for Products That Win by Steve Blank

Masters of Scale podcast with Reid Hoffman

Guest Bios 

Shub Agarwal is an associate professor of professional practice at the University of Southern California, an industry executive, and an advisor to start-ups and academic institutions. He holds an MBA from the University of California, Los Angeles (UCLA), and an MS from Carnegie Mellon University (CMU). He is the author of two books: Solve Catch-22 of Product Management and Successful AI Product Creation: A 9-Step Framework. He has made significant contributions to the fields of artificial intelligence and machine learning, holding several U.S. and global patents for his work, and is also a published author of several technical research papers.

With around two decades of extensive experience in product management and leadership, his journey has been marked by a relentless pursuit of leveraging AI technologies to create impactful products that redefine industry standards. His industry experience includes leadership roles at Amazon, Silicon Valley start-ups, and other Fortune 50 firms.

Hear more from Cindi Howson here. Sponsored by ThoughtSpot.

Learn what happens when the executive accountable for data strategy is also the executive accountable for the business results that depend on it. Saugata Saha, President of S&P Global Market Intelligence and Chief Enterprise Data Officer at S&P Global, shares how he manages one of the world's largest financial data estates while driving business outcomes across public and private markets. He breaks down the four pillars of S&P Global's data strategy, the federated organizational model that connects data teams to business value, and why capturing ROI from AI requires deliberate workflow transformation.

Key Moments

Why Data Strategy Must Follow Business Strategy (04:57): Saugata challenges the idea that data and business strategy can run in parallel. Market trends, customer pain points, and existing capabilities must come first.

Building an AI-Ready Financial Data Estate (15:10): Scale alone does not create intelligence. Saugata explains why semantic layers and graph databases are the hard work behind connected financial data.

How AI Compresses Post-Acquisition Data Integration (18:29): Manual reconciliation of millions of records is no longer the only path. Discover how AI entity matching accelerated post-acquisition integration.

The Federated Model That Connects Data to Value (22:49): Most large organizations either over-centralize data teams or leave them too embedded to scale. Saugata outlines the federated model that actually bridges both.

Rethinking AI Productivity: From Marginal to Transformative (28:29): Most AI programs stop at training and tooling. Saugata explains why deliberately redesigning workflows is the missing step between AI investment and real ROI.

Key Quotes

“Data strategy and business strategy have to be very tightly connected. And if they're not, that's when value capture does not happen. In fact, I would go so far as to say data strategy actually follows from business strategy.” - Saugata Saha

“Stop treating data as an afterthought or byproduct, but start thinking about data as a key ingredient for value creation and competitive advantage.” - Saugata Saha

“We don't want everybody to become 10% more productive, because that's a little squishy. We want 10% of the people to become a hundred percent more productive so they can do other things.” - Saugata Saha

“If a company can really use data at scale for better decision making, better client service, [and] better outcomes, that creates a lasting edge over the competition.” - Saugata Saha

Mentions

S&P Global Agrees to Acquire With Intelligence from Motive Partners for $1.8 Billion, Establishing Its Leadership in Private Markets Intelligence

The Data & AI Chief: Why a Federated Data Team is Crucial for Business Value, with Dow

Private Companies Wait Too Long to Go Public

The Lex Fridman Podcast

Guest Bios 

Saugata serves as President of S&P Global Market Intelligence, leading the division's efforts to deliver essential insights and intelligence to clients worldwide. He is also S&P Global’s Chief Enterprise Data Officer, responsible for driving innovation and excellence in the company’s enterprise data strategy. Saugata is a member of S&P Global’s Executive Leadership Team, contributing to the strategic direction and growth of the organization.

Before joining S&P Global, Saugata was a consultant at McKinsey & Company’s New York office, where he advised clients on strategy, mergers and acquisitions, corporate finance, and operational improvements across various industries, with a strong focus on financial services.

Hear more from Cindi Howson here. Sponsored by ThoughtSpot.

Learn why an organization’s ontology, a structured framework for how a business defines, connects, and makes sense of its data and knowledge, is the most valuable and most overlooked asset in any AI strategy. Jessica Talisman, CEO and Founder of The Ontology Pipeline, and Tony Seale, Founder of The Knowledge Graph Guys, break down what it actually takes to build trusted AI, covering everything from semantic layers and knowledge graphs to why provenance is non-negotiable. They explain how organizations can start building their knowledge infrastructure for AI, and make the case for why their ontology is their most defensible competitive asset.

Key Moments

BI Semantic Layers vs. AI Context Layers (02:21): Explore the evolution from 1990s business vocabularies to modern AI context layers. Learn why ontologies are essential for connecting data points beyond traditional BI.

Why Knowledge Graphs are Essential for AI (09:22): Understand why relational databases fail AI's needs. Tony explains how knowledge graphs turn data relationships into "first-class citizens" using open standards.

How to Build Your First Business Ontology (17:04): Stop over-modeling and start delivering. Learn how to anchor your data strategy to high-value use cases and business-language competency questions.

Solving the AI Provenance & Lineage Gap (34:19): Why LLMs lack built-in reliability. Jessica discusses the necessity of injecting data lineage at the retrieval layer to verify AI accuracy and prevent hallucinations.

Why Your Ontology is Your Most Valuable IP (39:27): In the age of commodity AI, your internal data relationships are your only moat. Discover why hosting your ontology with third parties puts your core assets at risk.

Key Quotes

“If you let somebody else take your ontology, learn the essence of what it is that you know that's out of distribution with the rest of the world, you've just given them everything valuable about your company.” - Tony Seale

“The accuracy of the information you receive is reliant upon the lineage or the provenance of the information received from an LLM. It's so important." - Jessica Talisman

“As a business leader, you need to be looking below the surface to the data infrastructure. The key trick to do right now is to turn the power of the models that we've got back upon your own internal infrastructure to build out these rich ontologies and to connect your information.” - Tony Seale

"Your ontology is like your thumbprint, your digital thumbprint for your organization. It's unique to each organization, and how you define things may not be the same as an LLM might define something." - Jessica Talisman

Mentions

The Ontology Pipeline® - A Semantic Knowledge Management Framework | Jessica Talisman

How the Ontology Pipeline Powers Semantic Knowledge Systems | Jessica Talisman

Why Early Knowledge Graph Adopters Will Win the AI Race | The Knowledge Graph Guys

Spec-First Development: Why LLMs Thrive on Structure, Not Vibes | The Knowledge Graph Guys

The Knowledge Graph Academy

Guest Bios 

Tony Seale

For over a decade, Tony has been passionate about linking data. His creative vision for integrating Large Language Models and Knowledge Graphs within large organisations has gained widespread attention, particularly through his popular weekly LinkedIn posts, earning him the reputation of ‘The Knowledge Graph Guy.’

Today, as the founder of The Knowledge Graph Guys, Tony is dedicated to helping organisations harness the power of their data. His consultancy develops cutting-edge Knowledge Graphs that fuel innovation and growth in the rapidly evolving Age of AI.

Jessica Talisman

Jessica Talisman has dedicated her 25-year career to exploring the dynamics of information and knowledge—how it flows across systems, evolves through context, and powers intelligent technologies. Her work spans historical research, educational frameworks, and enterprise-scale applications of artificial intelligence.

Previously a Senior Information Architect at Adobe, Jessica led the development of semantic knowledge graphs to enrich content and contextual understanding. She now serves as CEO and Founder of Ontology Pipeline, where she leads efforts to bridge the worlds of library science and data management - building robust, scalable knowledge systems for the AI era.

Hear more from Cindi Howson here. Sponsored by ThoughtSpot.

Discover how WHOOP is building an AI-powered health data infrastructure that is redefining how we understand human health. Emily Capodilupo, Senior Vice President of Research, Algorithms, and Data at WHOOP, explains how continuous physiological data is uncovering new opportunities in predictive health through AI, from presymptomatic disease detection to biological age scoring. She examines the governance challenges of deploying AI in a regulated environment and what it takes to build the data trust required to make it work at scale.

Key Moments:

How WHOOP Built Its AI and Data Foundation (00:57): Emily explains how WHOOP's early focus on elite athlete performance shaped the data collection rigor and multidisciplinary science organization that now powers its predictive health capabilities. She outlines the model she built across AI, machine learning, clinical research, and digital signal processing, and why starting with the highest-demand use case created a data foundation built to scale.

The Power of Continuous Data (06:21): Emily draws on WHOOP's sleep research to show how continuous physiological data reveals patterns that would be invisible without longitudinal tracking. She shares findings linking sleep architecture to metabolic disease, cancer risk, and cognitive decline, illustrating why the depth and continuity of a data set determine what insights are actually possible.

The Data Governance Challenge of Acting on Sensitive Data (13:17): Emily shares how WHOOP's respiratory rate data could detect COVID infection up to three days before symptom onset in over 80% of cases, but a denied FDA application left the company holding actionable insights it was legally prohibited from sharing. She examines the governance tension that emerges when your data capabilities move faster than the regulatory frameworks designed to govern them.

Turning Complex Multi-Signal Data Into a Single Actionable Metric (27:32): Emily introduces WHOOP's Healthspan feature, which translates physiological and behavioral data across nine components into a single biological age score tied to all-cause mortality risk. She explains why distilling complex data into one number is more motivating than presenting raw risk statistics, pointing to research that shows how age-based framing drives stronger behavior change.

Building Data Trust and Privacy Infrastructure at Scale (31:40): As WHOOP moves into FDA-cleared products and more sensitive data collection, Emily outlines the governance principles that underpin member trust. She argues that for any organization building on sensitive personal data, the asymmetry between earning trust and losing it should be a foundational design constraint.

Key Quotes:

"It takes 13 years to earn the trust and one mistake to lose it. And that kind of asymmetry is constantly top of mind." - Emily Capodilupo

"We were able to show that we could detect COVID up to three days before symptom onset in over 80% of cases." - Emily Capodilupo

“ WHOOP has been collecting data [for] over 12 years. We're working on a lot of new types of algorithms that are able to help people understand their bodies in ways that we might not have appreciated…even just a couple years ago.” - Emily Capodilupo

"One of the ways that AI has advanced the product... is this ability to chat with WHOOP in natural language, the way you might chat to a doctor or a trainer or a coach." - Emily Capodilupo

Mentions

Harvard Study | Analyzing changes in respiratory rate to predict the risk of COVID-19 infection 

Cornell Study Uses WHOOP Sleep Data to Monitor Patients at Risk for Alzheimer’s

Can Data Help Us Sleep Better? | WHOOP

There's More to Sleep than Sleep Need: The Importance of Sleep Consistency | WHOOP

Cribsheet & Expecting Better 2 Books Collection Set By Emily Oster 

The Family Firm: A Data-Driven Guide to Better Decision Making in the Early School Years By Emily Oster 

Guest Bio 

Emily Capodilupo is an award-winning AI and research leader with more than 13 years of experience building and scaling science-driven organizations in fast-paced startup environments. She began her career as an emergency medical technician before studying neurobiology and human sleep at Harvard University and conducting research at Brigham and Women’s Hospital. Emily is driven by a passion for using data to solve hard problems and advance our understanding of human physiology. Along the way, she "accidentally" became a data scientist, recognizing that the biggest breakthroughs in health require not just rigorous science, but big data and bold technology. 

As WHOOP’s first employee, Emily founded and now leads the company’s science organization, pioneering a new model of health that begins long before diagnosable illness and is continuous, personalized, AI-powered, and designed to empower individuals to take the driver’s seat in their own well-being. She has built and scaled multidisciplinary teams across artificial intelligence, machine learning, digital signal processing, clinical research, and engineering to translate real-time physiological data into actionable insights that improve performance, resilience, and long-term health. Emily’s work sits at the intersection of wearable technology, digital biomarkers, and predictive health, helping shift healthcare from reactive treatment to proactive optimization.

Hear more from Cindi Howson here. Sponsored by ThoughtSpot.

Learn why scaling AI is as much a human challenge as it is a technological one. Stefano Puntoni, Co-Director of Wharton Human-AI Research and Professor at The Wharton School, examines the limits of data-driven decision making in the age of AI and why insights so often fail to translate into action. He breaks down the psychology behind AI resistance and outlines the leadership and change management strategies needed to turn AI potential into real organizational impact.

Key Moments:

Why More Data Doesn’t Lead to Better Decisions (02:26): Stefano challenges the assumption that smarter algorithms automatically produce smarter decisions. He argues that decision quality depends on rigorous conceptual thinking before turning to data. Without clearly defining objectives, alternatives, and success criteria, analytics efforts rarely translate into meaningful action.

Conversational AI and the Lowering of the Cost of Action (07:26): Stefano explains how conversational AI brings decision makers closer to data by reducing friction. By lowering the cost of experimentation, AI enables managers to test hypotheses in real time instead of waiting days for analysis. This shift moves organizations from analysis paralysis to faster, more confident action.

Rethinking Your Role in the Age of AI (17:16): For professionals navigating disruption, Stefano outlines two paths forward. One is becoming a complement to AI by upskilling and using the technology as a productivity multiplier. The other is pivoting toward skills AI is less likely to replace, such as strategy, orchestration, and human judgment.

The AWARE Framework: Pairing Technical Rollout with Human Rollout (22:41): Stefano introduces the AWARE framework to help leaders anticipate and manage the human reactions to AI transformation. He argues that every technical implementation must be matched with structured communication, identity support, and organizational alignment. Without this dual-track approach, even well-designed AI systems can fail to gain traction.

Change Management, AI Literacy, and the Gap in Organizational Readiness (31:11): Only a small percentage of organizations have formal AI change management programs. Stefano questions whether companies are truly prepared for large-scale AI transformation. He emphasizes that AI literacy, leadership accountability, and structured change management will determine whether AI investments translate into sustained performance.

Key Quotes:

“ The leaders need to know why we are doing AI. AI is not a strategy; AI is just a tool. So what is it that we're trying to achieve?” - Stefano Puntoni

“ I think the problem is that technology is almost like taking all the oxygen from the room. There's so much attention and urgency around the tech itself that we often forget the people around it.” - Stefano Puntoni

“You don't want to be the substitute to the technology because if that is what you do, then there's no future. But if you're a complement, the technology might be a multiplier of your productivity.” - Stefano Puntoni

Mentions

Decision-Driven Analytics: Leveraging Human Intelligence to Unlock the Power of Data

The Wall Street Journal: The Boss Has a Message: Use AI or You’re Fired

2025 Report Accountable Acceleration: Gen AI Fast-Tracks Into the Enterprise

How AI Affects Our Sense of Self

Why Gen AI Feels So Threatening to Workers

Conversational AI: The Next Frontier of Digital Platform Monetization

Guest Bio 

Stefano Puntoni is the Sebastian S. Kresge Professor of Marketing at The Wharton School. Prior to joining Penn, Stefano was a professor of marketing and head of department at the Rotterdam School of Management, Erasmus University, in the Netherlands. He holds a PhD in marketing from London Business School and a degree in Statistics and Economics from the University of Padova, in his native Italy.

His research has appeared in several leading journals, including Journal of Consumer Research, Journal of Marketing Research, Journal of Marketing, Nature Human Behavior, and Management Science. He also writes regularly for managerial outlets such as Harvard Business Review and MIT Sloan Management Review. Most of his ongoing research investigates how new technology is changing consumption and society, including how humans are adopting and evolving with AI.

Hear more from Cindi Howson here. Sponsored by ThoughtSpot.