Guest:
Sujay Dutta, Author - Data as the Fourth Pillar | Global Account Lead | Databricks
Siddharth Rajagopal, Author - Data as the Fourth Pillar | Chief Architect at Informatica | Executive MBA at Quantic
Host: Seth Earley, CEO at Earley Information Science
Published on: January 19, 2026
This episode welcomes Sujay Dutta and Siddharth Ragagopal, co-authors of Data as the Fourth Pillar. With extensive experience guiding global organizations on aligning data strategy with real-world business outcomes, Sujay (based in Stockholm) and Siddharth (based in the Netherlands) offer deep insights into AI adoption, data governance, and scaling artificial intelligence responsibly. Hosted by Seth Earley, the conversation explores how businesses can move beyond AI experimentation and develop a mature, impactful data strategy.
Key Takeaways:
Featured Quote from the Show:
"One of the key challenges with AI is not about AI being ready for people, but are people ready for AI? ... Ultimately it will land upon the people of the enterprise. How the leaders are clarifying that incentive model to each individual." — Sujay Dutta
Tune in to learn how to build a solid data foundation, avoid common AI pitfalls, and prepare your organization—and your people—for the future of intelligent business.
Links
LinkedIn: https://www.linkedin.com/in/sujaydutta
LinkedIn: https://www.linkedin.com/in/sidd-rajagopal/
Website: https://datathefourthpillar.com
Ways to Tune In:
Earley AI Podcast: https://www.earley.com/earley-ai-podcast-home
Apple Podcast: https://podcasts.apple.com/podcast/id1586654770
Spotify: https://open.spotify.com/show/5nkcZvVYjHHj6wtBABqLbE?si=73cd5d5fc89f4781
iHeart Radio: https://www.iheart.com/podcast/269-earley-ai-podcast-87108370/
Stitcher: https://www.stitcher.com/show/earley-ai-podcast
Amazon Music: https://music.amazon.com/podcasts/18524b67-09cf-433f-82db-07b6213ad3ba/earley-ai-podcast
Buzzsprout: https://earleyai.buzzsprout.com/
Podcast Transcript: Data as the Foundation for AI Success
Transcript introduction
This transcript captures a conversation between Seth Earley, Sujay Dutta, and Siddharth Rajagopal about why data readiness is the critical prerequisite for AI adoption. The discussion covers how organizations can move beyond pilot programs, how to align business demand with technical supply capabilities, and why capturing tacit human knowledge is the defining challenge of enterprise AI maturity.
Transcript
Seth Earley: Welcome to the Earley AI Podcast. I'm your host, Seth Earley. Each episode explores how artificial intelligence is transforming the way organizations manage information, make decisions, and create better outcomes for customers and employees. Today we are focusing on the role of data as the foundation for every successful AI initiative. Joining me are two guests who spent their careers helping organizations align data strategy and real business outcomes. I'm pleased to welcome Sujay Dutta and Siddharth Rajagopal, authors of Data as the Fourth Pillar. Together, they bring deep experience guiding leaders through the realities of AI adoption, from preparing data and governance to scaling AI responsibly in complex environments. Sujay, Sidd, welcome to the show.
Sujay Dutta: Thank you for having me.
Siddharth Rajagopal: Thanks for having me.
Seth Earley: Where are you guys based?
Sujay Dutta: I am in Sweden, Stockholm.
Siddharth Rajagopal: And I'm in the Netherlands, so we are both in Europe.
Seth Earley: I'm in Boston, and there's a ton of snow outside right now. So let's start off with misconceptions about AI, always a good place to start. What are the things that you see in your world, especially among leaders trying to decide where to invest and how to get started?
Sujay Dutta: AI is being misconstrued as just a technology. In my view, AI impacts every aspect of your business, the people, the processes that need to be updated and changed, the technology, and also whether you're prepared for the data. There's a holistic angle to it, and it's just not a technology play. Technology is just one of the factors.
Siddharth Rajagopal: I'd add that a lot of business leaders see generative AI, things like ChatGPT or Gemini, and consider that the entire spectrum of AI. But the spectrum of AI is really wide. First of all, every business problem doesn't need an AI solution. And within that, every AI solution doesn't require a generative AI solution. Gen AI is very good at specific things, generating text, summarization, content creation, but may not be suited for, say, a sales forecast model, where a machine learning approach might be much more cost-effective and accurate. It's understanding the nuances of what exactly is possible.
Seth Earley: In terms of how organizations decide where to invest and how to get started, a lot of them are doing multiple pilots and initiatives simultaneously. How do they step back and put a clear evaluation framework in place?
Siddharth Rajagopal: The first thing is to understand what you need to do as an enterprise with AI, that's really a board and C-level decision. Once you have a good understanding of what is possible and what you want to achieve, you need to make sure your people, your process, and your data are ready. A lot of enterprises dive straight into AI and later find out they're not ready, either the people who use and build these systems don't know how to do it most effectively, or the data isn't ready. Generative AI models are essentially commodity black box models. Unless you provide the right context, the right data, metadata, enterprise tone, and information about what your organization can and cannot do, your AI model won't succeed. For leaders, I would say: gauge your data maturity before looking at your AI ambitions.
Sujay Dutta: To add to that, boards and CEOs are pressured to show AI outcomes, so some level of balancing act is required. Experimentation will happen, and that's the nature of innovation. You will fall, learn from mistakes, and the journey is never a straight upward curve. But beginning the journey is important, and it's not a one-size-fits-all approach. A large multinational conglomerate faces very different challenges than a startup. Every company has to decide for themselves based on the capital and resources they have.
Seth Earley: In your book, you talk about the challenges of getting stuck in pilot mode. What patterns do you see in organizations that struggle to move beyond experimentation, and what differentiates those that successfully scale?
Sujay Dutta: When companies plan experimentation as part of a journey, that's when it works. Leaders higher up the organization chain have visibility into strategic goals, and they plan how AI is going to impact their two-, three-, and five-year goals. Having that telescopic journey view is critical. It's like going on a hike, you plan your day, but you execute on a shorter time basis, three to six months at a time.
Seth Earley: Having that North Star in mind, but getting there incrementally, learning from pilots, curating data, understanding use cases, clarifying ownership. How do you work with leaders to define the right starting points and identify measures that reflect business value? Organizations often start at a high level with ambiguous goals, increase revenue, improve customer service, and it needs to break down into very specific objectives and process steps. How do you see organizations getting from arm-waving board-level goals to the real operational detail?
Siddharth Rajagopal: One pattern I see is looking at AI as a separate thing rather than embedding it as part of your entire business process. Large enterprises already have an operating model for people, processes, and technology. When you have a big business ambition. say, opening a new plant in Mexico, you already have a process for setting up people structure, complying with local laws, and selecting the right technologies. Why not extend that same framework to embed data and AI as well? Rather than just handing it to IT and hoping for magic results. The second pattern is related to data estate. When your data is siloed, like in the Verizon example Seth mentioned, the order data never flows through to execution systems. Data plays an important role in bridging the gap between what boards decide and what actually happens in operations. It's not just data for AI, it's AI for data too. Both need to be woven into every aspect of business operations.
Seth Earley: And the Audi example from your book, can you walk us through that?
Siddharth Rajagopal: Audi's CDO found that when they analyzed challenges across their digital transformation, business pressure appeared low as a challenge. But when they double-clicked on why, they found it wasn't because everything was fine, it was because not enough business demand was coming in. The business teams were building solutions on their own and the IT teams were doing their digital transformation on their own. They weren't connected. A great specific example was the shop floor welding process. Welds are done continuously, and if a bad weld is only caught at a later quality control stage, it needs to be reworked at significant cost. Initially, they looked at using computer vision AI, taking pictures and classifying good versus bad welds. But the data leader came in and identified that temperature data was an additional input that could significantly improve model accuracy. That required business understanding, engineering know-how, and a data and AI leader working together. It's what's missing in large enterprises where people still operate in silos, when they come together, they find much larger solutions to problems.
Seth Earley: So the root cause is often that teams don't know what they don't know, they don't understand what's possible, which is why there's less business demand. The answer to that is education around possibilities and ideation, bringing cross-functional teams together to analyze problems systematically.
Siddharth Rajagopal: Exactly.
Sujay Dutta: And that's where we formulated the maturity model framework in our book. The classic gap in organizations is between demand and supply. Like economics 101, you cannot be successful until demand and supply meet. Once you document all your business problems and challenges as use cases — that's your demand. On the other side, you build out your supply capabilities: AI technology, model training, data platforms. In a classic organization, these are far separated, which is exactly what happened at Audi. The supply teams had no visibility of the demand, and the business teams had no visibility of what was happening on the supply side. That gap needs to be closed through a structured maturity framework, which is why we call data the fourth pillar.
Seth Earley: Can you walk us through the three stages of maturity?
Sujay Dutta: The three stages are: foundational, scale, and automated.
Siddharth Rajagopal: At the foundational stage, business and IT teams operate in silos, delivering data and AI value in pockets but not holistically. They may have good reporting or data platforms but are missing things like an ontology layer, a layer that allows them to understand what data means from different parts of the business. From a demand standpoint, different business units may be trying to solve problems independently, creating shadow IT. At the scale stage, you've enabled self-service of data and AI for business users who can leverage capabilities on their own. At the automated stage, you have most use cases with high quality, compliance, and speed demands met, and you've maximized your AI capabilities, deploying AI agents to run a truly autonomous enterprise.
Seth Earley: You mentioned quality, compliance, and speed as demand dimensions, can you explain how those are measured?
Siddharth Rajagopal: We use a framework abbreviated as QCS, quality, compliance, and speed. These are the dimensions we use to score the demand intensity of each use case. For quality, we use the standard six dimensions: accuracy, completeness, consistency, timeliness, validity, and uniqueness. The key is that these scores come from what the use case requires, not what you're currently supplying. For example, a fraud detection engine may require real-time data timeliness, while a marketing campaign may tolerate data that's a few days old. You score your use cases on these dimensions, and that positions you on the demand axis. Combined with an assessment of your supply capabilities, that places you in the maturity matrix, from foundational in the lower left, to automated in the upper right.
Seth Earley: What about the human dimension, skills, roles, change management? What guidance do you have for leaders feeling overwhelmed about preparing their teams?
Siddharth Rajagopal: First, identify what your enterprise structure already is, centralized, decentralized, or federated, and extend that same structure to your data and AI function. We see situations where IT is centralized but the entire business is decentralized, making it very difficult to map capabilities across. Then, look at the maturity of each business domain. Some may need significant hand-holding from an IT standpoint, while others prefer to operate more autonomously and consume data as a service.
Sujay Dutta: And you must follow your organization's culture and ways of working. You cannot create a data and AI structure that is totally disjointed from how the organization actually operates — that's a recipe for failure. A lot of study needs to be done to understand the culture, the incentive model, and the operating reality before you get to solutioning.
Seth Earley: A recurring theme in your work is that AI increases the importance of clean, structured, well-governed data. What core data and knowledge practices are essential for sustainable AI adoption?
Siddharth Rajagopal: A lot of data teams are well-equipped technically, data engineers, data scientists, but there aren't enough people in data organizations who deeply understand the business and are focused on adoption. You need people who can function almost like business development or sales roles within a data organization, people who can sell, in a sense, the value of the data landscape to business units and drive adoption. That mix of technical expertise with business acumen and adoption focus is what's often missing.
Seth Earley: Where do you see the most meaningful opportunities for AI over the next couple of years, and what steps should leaders take to prepare?
Sujay Dutta: The key challenge with AI is not about whether AI is ready for people, it's whether people are ready for AI. The easy wins are individual productivity gains: using Gen AI to summarize emails, understand documents. But to get to the automated stage and truly autonomous AI agents, the tacit knowledge of individuals needs to be captured, the business context that isn't properly documented today. And it's not just about documentation processes. There's an incentive problem. Someone who has been in an organization for 10 or 15 years, their differentiation is the knowledge they hold. Why would they document everything for an AI agent to consume, if that makes them less relevant? That is the key roadblock to business value realization beyond individual productivity. It's not a technology problem. It will ultimately land on how leaders clarify the incentive model to each individual.
Siddharth Rajagopal: And from the C-suite standpoint, it's about reimagining your business, not just embedding AI into existing processes, but genuinely challenging those processes and asking what they would look like if designed from scratch with AI. Leaders need to look two or three steps ahead, not just implement what's available today. ChatGPT came three years ago, how many of us are still using that base model? The pace of change demands proactive, forward-looking leadership.
Seth Earley: Great. Well, Sujay and Sidd, thank you very much for joining our conversation today and sharing your insights. For our listeners, we'll include links to Data as the Fourth Pillar and related resources in the show notes. Thank you all for listening, and we will see you next time on the Earley AI Podcast.
Siddharth Rajagopal: Pleasure is all ours, Seth. Thank you.
Sujay Dutta: Lovely discussion. We truly enjoyed it.