Patrick Van Deven: The Frontier Firm Has a Data Problem

Patrick Van Deven: The Frontier Firm Has a Data Problem In this episode of Scouting for Growth, Sabine VanderLinden sits down with Patrick Van Deven to unpack one of the biggest hidden blockers to becoming a true AI-native enterprise: legacy data infrastructure. As organizations rush toward the “Frontier Firm” vision championed by Microsoft — intelligence on tap, human-agent collaboration, and AI-powered workflows — Patrick argues that most regulated industries are still running on fragmented data pipelines built decades ago. Beneath the excitement around agentic AI lies a critical operational reality: data remains horizontally distributed across systems such as SAP, Salesforce, Guidewire, and legacy warehouses, stitched together by opaque code that no one fully understands anymore. Patrick explains why the future of AI in regulated industries depends less on flashy copilots and more on deterministic, governed, audit-ready data transformation. Drawing from his 35 years in enterprise software and his leadership at Volspeed, he outlines how AI is now reshaping data engineering itself — automating the “plumbing” layer while generating the metadata and lineage AI systems need to operate responsibly. Together, Sabine and Patrick explore why re-architecting does not require a dangerous core system replacement, how organizations can solve tractable business problems in months rather than years, and why the next generation of enterprise leaders must bridge business expertise and data intelligence. This conversation is a practical roadmap for any executive navigating AI transformation inside complex, regulated environments.   KEY TAKEAWAYS What stood out most to me in this conversation with Patrick was the reality that the “Frontier Firm” conversation is no longer about experimentation. It is about operational readiness. Every organization I speak to wants intelligence on tap, agentic workflows, and AI-enabled productivity, yet many are still constrained by fragmented legacy systems and undocumented data logic buried deep inside their infrastructure. Patrick made it very clear: if we do not solve the data foundation problem, we simply accelerate complexity and risk. One insight that resonated deeply was the idea that data engineering is entering the same transformation that software engineering experienced with generative AI. The real opportunity is not just automation, but abstraction — enabling smaller teams to solve historically impossible integration problems while creating governed, machine-readable metadata that AI systems can actually trust and consume responsibly. I was also struck by Patrick’s perspective on talent. Rather than replacing expertise, AI elevates the importance of subject matter experts who understand the business context behind the data. The future belongs to professionals who can bridge operational understanding with technical fluency and collaborate effectively with AI-enabled systems. Most importantly, this conversation reinforced that becoming a Frontier Firm does not require ripping out every core system overnight. The no-regret move is to start solving tractable, high-value data problems now — especially those tied to governance, lineage, regulatory reporting, and customer intelligence. Organizations that modernize their deterministic data layer today will be the ones capable of building scalable, trustworthy AI tomorrow.   BEST MOMENTS “You can bolt all the AI you want on top of that. It will not make you a frontier firm. It will just make your regulatory problems arrive faster.” — Sabine VanderLinden “AI is coming to data engineering just like it came to software engineering.” — Patrick Van Deven “The board looks at AI at the end of the value chain of data. But how did that data come to be?” — Patrick Van Deven “There is no world where a company would run on one system.” — Patrick Van Deven “Treat the AI agent like an employee. Onboard it, brief it, give it a personality.” — Sabine VanderLinden “The dragon in the basement has finally reached the boardroom.” — Patrick Van Deven “No data lineage. No agent bosses. No governed transformation. No intelligence on tap.” — Sabine VanderLinden “This is a new era for subject matter experts.” — Patrick Van Deven   ABOUT THE GUEST Patrick Van Deven is the CEO of Vaultspeed and a veteran enterprise software leader with more than 35 years of experience in software engineering, predictive analytics, data infrastructure, and venture investing. Patrick began his career as a software engineer, building and selling his first commercial application at just 22 years old. He later spent 15 years at SAS Institute, where he helped build data and predictive analytics applications for enterprise environments. He then transitioned into venture capital as an Operating Partner and General Partner at Fortino Capital, investing in software and AI startups across Europe. In 2025, Patrick stepped back into an operational leadership role as CEO of Vaultspeed, driven by his belief that automating deterministic, governed data transformation is one of the most critical “no-regret moves” organizations can make in the age of AI. Today, Vaultspeed works with major global enterprises, including organizations operating across highly regulated industries such as insurance, banking, and financial services.   ABOUT THE HOST Sabine VanderLinden is a corporate strategist turned entrepreneur and the CEO of Alchemy Crew Ventures. She leads venture-client labs that help Fortune 500 companies adopt and scale cutting-edge technologies from global tech ventures. A builder of accelerators, investor, and co-editor of the bestseller The INSURTECH Book, Sabine is known for asking the uncomfortable questions—about AI governance, risk, and trust. On Scouting for Growth, she decodes how real growth happens—where capital, collaboration, and courage meet. If this episode sparked your thinking, follow Sabine VanderLinden on LinkedIn, Twitter, and Instagram for more insights. And if you’re interested in sponsoring the podcast, reach out to the team at [email protected]