The Hidden Crisis Lurking Behind Every Hospital Software Upgrade
When a health system announces a shiny new electronic health record platform, the headlines celebrate modernization, efficiency, and better patient care. What nobody photographs is the decades of medical history left stranded in the old system—the diagnoses, medications, allergies, and clinical notes that suddenly become harder to reach than buried treasure. On this episode of Business Unmasked, host Anupama sits down with Dominique Gross to pull back the curtain on one of healthcare's most consequential yet least visible challenges.
Gross has built his career around a question most hospital executives would rather not confront: What actually happens to patient data when the institution upgrades its digital backbone? The answer, he explains, involves far more than a simple technical migration. Years of records must be moved, stored in accessible formats, and preserved in ways that allow clinicians to retrieve critical information at moments when seconds matter. The work is unglamorous, expensive, and absolutely foundational to both daily care and the future of medicine itself.
The infrastructure for moving and storing healthcare data isn't just backend plumbing—it's the foundation that determines whether AI can ever be safely adopted in clinical settings.
From Engineer to CEO: Reframing Technical Value
Gross's path to the corner office defies the typical Silicon Valley script. He started in engineering and marketing, disciplines that taught him to build products and communicate their worth, but neither fully prepared him for the CEO's chair. That transition required a different kind of rewiring: learning to see the company itself as a product needing constant iteration, and understanding that growth demands rebuilding foundations rather than simply adding features.
When he took the helm of a decade-old company, Gross faced a familiar trap. Ten years of operation had created legacy systems—not in code alone, but in processes, culture, and assumptions about what customers valued. Rebuilding for growth meant dismantling what had worked well enough in the past and reconstructing around where the market was heading. For healthcare technology specifically, that direction points toward interoperability, cloud infrastructure, and the looming imperative of AI readiness.
Perhaps his most hard-won insight bridges the technical and commercial worlds. Gross argues that technical products—especially in complex domains like healthcare IT—fail when sellers lead with specifications and capabilities. The breakthrough comes from framing every conversation around the problem the technology solves. Hospital administrators drowning in data-retrieval costs and compliance risks don't wake up wanting middleware; they want assurance that their clinicians won't miss a critical allergy warning because it lives in a retired system from 2014.
Technical products sell better when you stop talking about what they do and start talking about the problem they eliminate for the person staring at the purchase order.
Why Data Legacy Matters for AI's Future in Medicine
The episode's most urgent thread connects yesterday's data archival to tomorrow's clinical AI. Gross makes clear that the current frenzy around artificial intelligence in healthcare rests on assumptions about data quality and accessibility that remain largely unexamined. An algorithm is only as trustworthy as the information feeding it, and that information must be complete, accurately migrated, and properly contextualized across systems that were never designed to speak with one another.
Without the painstaking work of preserving historical patient records in usable form, AI tools risk training on incomplete pictures of human health—or worse, missing entirely the longitudinal patterns that matter most for predictive care. The "safer AI adoption" Gross emphasizes depends on infrastructure investments that offer no immediate ROI and generate no patient-facing buzz. They are the kind of expenditures that get deferred until a crisis exposes the gaps.
Key Takeaways for Founders
1. The invisible infrastructure determines your ceiling. Whether in healthcare or any regulated industry, the work customers never see—data architecture, compliance plumbing, legacy integration—creates the conditions for everything that follows. Underinvest here and every ambitious feature becomes structurally precarious.
2. Rebuilding beats patching for sustainable growth. Gross's experience transforming a ten-year-old company underscores that incremental fixes on aging foundations eventually collapse under scaling pressure. Founders must recognize when evolutionary improvement has reached its limit.
3. Problem-first framing outperforms feature lists. Even deeply technical products find their market when sellers articulate the specific pain they eliminate rather than the capabilities they deploy. This translation skill separates enduring companies from those that stall in pilot purgatory.
4. AI readiness is a data readiness problem. Before adopting artificial intelligence tools, organizations must audit whether their historical information is accessible, accurate, and complete. The safety and effectiveness of algorithmic healthcare depends entirely on this unglamorous precondition.
