Nov 5, 2025 · 27m · big-technology
Why Getting Data Right Could Be The Key To Effective AI Projects — With Charles Sansbury
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Cloudera CEO Charles Sansbury joins Alex Kantrowitz to examine the economic and operational realities of enterprise AI, explaining why high-fidelity proprietary data, hybrid orchestration architectures, and rigorous governance are essential to overcoming ROI hurdles.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Alex holds 19.7% of the talking time here. How this is scored →
Alex as informed peer 5.3 Guest teaching 3.9 Guest disagreement 1.1 Alex pushing back 2.1
speaking balance: gold is Alex, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 7:25 Firm defense of IT governance necessity
Sansbury firmly rejects the premise that business units should bypass IT oversight, stating personal discomfort with releasing uncontrolled autonomous agents into enterprise infrastructure.
Hardest push from Alex ▶ 19:02 Pushing the CEO dilemma between core cash cow and AI reinventionKantrowitz directly challenges Sansbury on whether Cloudera is clinging to legacy non-AI data platforms instead of embracing a true Day 1 reinvention.
Biggest teaching moment ▶ 1:23 Explaining the economics of dot-com speculation versus modern infrastructureSansbury educates the host on the mathematical logic of 1990s venture bets like Webvan compared to the massive hardware and electricity expenditures driving modern AI.
Alex holds their own ▶ 4:07 Synthesizing dot-com roulette against centralized AI adoptionKantrowitz demonstrates sharp analytical synthesis by contrasting the fragmented dot-com venture model with today's massive capital concentration in actual utilized platforms like OpenAI.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Alex as informed peer | Guest teaching | Guest disagreement | Alex pushing back | Why |
|---|---|---|---|---|---|---|
| Comparing the Dot-Com Bubble to Modern Artificial Intelligence Investments | 6 | 4 | 1 | 1 | Alex Kantrowitz establishes his familiarity with Charles Sansbury's background in dot-com era investment banking and software. Sansbury explains the structural and financial differences between 1999 venture capital dispersion and modern AI capital concentration, which Kantrowitz synthesizes effectively. | |
| Navigating Grassroots AI Adoption and Enterprise IT Governance | 5 | 4 | 2 | 3 | Kantrowitz references Cloudera's state of enterprise AI survey to question whether business units like marketing really need IT governance. Sansbury defends strict oversight, explaining that governance ultimately acts as an accelerant rather than a bottleneck when deploying autonomous agents. | |
| Solving Enterprise Data Fragmentation with Modern Orchestration Architectures | 6 | 5 | 1 | 2 | Kantrowitz highlights the survey data showing only nine percent of enterprises have all data accessible. Sansbury details the enterprise reality of fragmented data silos and explains Cloudera's hybrid data orchestration strategy. | |
| Deploying Agentic AI for Anti-Fraud in Banking and Private Data | 6 | 5 | 1 | 3 | Sansbury illustrates enterprise AI value through a detailed case study of a global bank automating anti-fraud investigation workflows. Kantrowitz presses on where human oversight lines are drawn and how proprietary data security is preserved. | |
| Reconciling AI ROI Deficits with Cloudera's Dual Business Strategy | 7 | 4 | 2 | 4 | Kantrowitz challenges Sansbury on the MIT study reporting a 95 percent lack of AI ROI and presses him on whether Cloudera should disrupt its billion-dollar core business. Sansbury defends a dual strategy of protecting foundational data platform revenue while acquiring modern AI capabilities. | |
| Maximizing Model Performance Through Data Lineage and Domain Specialization | 6 | 5 | 1 | 2 | Kantrowitz discusses enterprise hallucination risks and car dealership chatbot failures. Sansbury explains that as foundational model parameter scaling plateaus, proprietary domain context and data lineage become the primary competitive differentiators. | |
| Conclusion and Cloudera AI Resource Recommendations | 1 | 0 | 0 | 0 | Brief standard outro where Sansbury directs listeners to Cloudera resources and Kantrowitz wraps the interview. |
Statements from this episode (10)
Sansbury: AI gains will outpace losses via tiny group of massive winners
“Certainly, there's a lot of money that's going to be lost in these investments. But I think what's going to happen is actually different, in that I think there's going to be, you know, a lot of companies that turn out not to be successful, but a very small num…”
Cloudera CEO: 96% of companies try AI, but IT approves only 30%.
“If 96% of companies are trying AI, only about 30% of IT organizations have approved of what they're doing.”
Sansbury: Autonomous AI agents operating without humans create unmanaged enterprise risks
“The concern, especially as you move from generative to agentic AI, where you have autonomous agents moving through your systems doing stuff without checking back in with a human, I think it creates risks that we haven't got our hands around yet.”
Cloudera Study: Only 9% of Enterprises Have All Data Accessible for AI
“Just nine percent of respondents sent that all of their data was available. And only 38% said that most of the organization's data was available.”
Sansbury: Moving All Data to Cloud Sacrifices Proprietary Enterprise Context
“What our perspective has been, the answer can't be you take all that data and move it to the cloud so it can run very neatly on these cloud-based models, because then you lose kind of control over that enterprise context that you built over years, the transact…”
Sansbury says an AI agent replaced a 1,000-person banking fraud team.
“And it's allowed them to basically take a team of a thousand and repurpose the majority of those people to other functions within the bank. So that's a savings of tens of millions of dollars and a more efficient process and better for the customer who gets his…”
Sansbury: Threshold for Human Intervention in AI Workflows Will Keep Rising
“I will tell you if the improvement in the large language models that I use is an indication, the line's going to keep going up. The technology is iterating very fast and getting better much more quickly than I would have thought.”
Sansbury: Enterprise AI failures stem from disconnect between IT and business users
“A lot of the use cases I believe are being driven by, you know, business users who don't have as much technology experience, or IT users, Who don't have as much business experience driven by the urgency of, oh my gosh, we got to do something.”
Sansbury: Enterprise AI Value Requires Proprietary Context Over General Pre-Training
“It's only valuable to train on our own internal context, A, and B, that model does not have to have read War and Peace to deliver to us that it doesn't have to be generally trained on everything that's available. It has to be trained on our content. And so the…”
Sansbury suggests base AI models are hitting an asymptotic performance plateau.
“Models are getting better, but maybe they're approaching this kind of asymptotic barrier where are thirteen billion parameters really that much better than twelve billion parameters, right? Maybe nine is good. Maybe five. We, someone mathematically has done th…”