Healthcare systems and insurance companies are entering a new era of artificial intelligence implementation, but given how fast technology and markets move, purchasing strategies have yet to catch up. Many people are buying AI the same way they bought software 10 years ago. That means one tool, one use case, one department at a time.
The result of today’s rapidly changing market is a hodgepodge of vendors, pilot purgatory, increased integration complexity, and uncertain return on investment.
Moving from experimentation to enterprise-wide deployment requires healthcare leaders to rethink how they evaluate, procure, and manage AI across their organizations. This change requires moving beyond point assessment to intentional system design. Increasingly, buyers are not asking if a solution works, but instead asking where it fits, what it depends on, and how it will add value over time.
AI can no longer be treated as an application decision. It needs to be architectural. Leaders need to stop evaluating tools in isolation and start designing a coherent enterprise AI architecture.
Organizations have the potential to succeed through a new framework for AI procurement. The framework recognizes three distinct layers: a core enterprise platform, a foundational model platform, and specialized context-aware use case innovators. For startups, this architectural view is no longer an option when considering go-to-market strategies. This is rapidly becoming the lens through which business buyers assess relevance, durability, moat, and long-term strategic fit.
Layer 1: Core Infrastructure and Enterprise Platform
Core infrastructure and cloud platforms now form the foundation of healthcare enterprise stacks led by hyperscalers such as Microsoft Azure, AWS, and Google Cloud. These providers provide the compute, storage, security building blocks, and data services that power modern AI capabilities. On top of this infrastructure layer have been systems of records and enterprise application vendors (particularly Epic and platforms like Workday, ServiceNow, and Salesforce) that have long sat and continue to serve as the digital backbone for most healthcare organizations.
These vendors are increasingly working together to embed AI directly into the enterprise stack, leveraging privileged access to the data, workflow, identity, and compliance frameworks that health systems and payers already trust. Hyperscalers in particular are becoming foundational not only as infrastructure providers but also as policy setters that define how data is stored, moved, secured, and increasingly large-scale AI workloads are managed.
For startups, this layer defines the company’s center of gravity. Core platforms and cloud providers shape procurement constraints, deployment patterns, and architectural boundaries long before point solutions are evaluated. They determine where your data resides, how your models are accessed, and what “enterprise-ready” actually means. While this layer provides the lowest-risk implementation path for buyers, it compresses opportunities for vendors whose value propositions overlap too closely with native capabilities or cloud-embedded services. Startups that fail to account for this reality risk being sidelined by evolving platforms rather than being supplanted by direct competitors.
Layer 2: Foundation Model Platform
The emergence of foundational model platforms presents both opportunities and strategic tensions. New products like OpenAI’s Frontier and Anthropic’s Cowork point to a future where enterprises can operate secure, organization-specific AI environments between core systems and end-user applications. And these products are increasingly being incorporated into Layer 1 infrastructure (for example, Microsoft’s recent announcement of Copilot Cowork, powered by Anthropic’s Cowork product).
These platforms act as an orchestration layer, enabling custom application development, centralized governance, reusable prompts and workflows, and controlled access to proprietary data. For startups, this layer is increasingly likely to become the control plane for enterprise AI. Products that align with it, rather than trying to circumvent it, are likely to benefit from shorter procurement cycles, deeper embedding into the enterprise, and a clearer path to expansion.
At the same time, the underlying platform poses existential questions. If these become the default operating system for enterprise AI, startups will need to decide whether to build on top of them, alongside them, or in competition with them. The choices you make will impact your product architecture, deployment model, data ownership, and even pricing strategy. Startups that fail to clearly articulate this positioning may struggle to gain corporate trust, regardless of model performance.
Layer 3: AI-focused startups
The world of AI startups tackling discrete operational pain points, from pre-authorization and claims management to revenue cycle optimization and clinical documentation, continues to expand. Many of these companies deliver quick, measurable ROI and solve problems that core vendors are slow to address.
But the hurdles are rising. Enterprise buyers are increasingly evaluating Layer 3 solutions not only for functionality but also for architectural compatibility. How does the product integrate with the core platform? Does it leverage an underlying model environment already approved by the organization? Can it scale without introducing fragmentation, redundant infrastructure, or governance risks?
As Layer 1 and Layer 2 mature, successful startups will be those that position themselves as complementary to the architecture rather than as independent tools. Aligning integration strategies, data boundaries, security posture, and governance are becoming as important as algorithmic sophistication.
strategic imperative
Procuring healthcare AI is no longer about picking winners one use case at a time. It’s about orchestrating across layers by integrating enterprise platforms, basic functionality, and specialized applications into a coherent system that can evolve.
This change requires a new level of entrepreneurial awareness from startups. Winning vendors not only understand the buyer’s problems, but also the buyer’s stack. They design products that integrate well into existing and emerging enterprise architectures and communicate with clarity and discipline.
The digital frontier of healthcare is no longer a matter of whether we adopt AI, but rather how we organize it. Sourcing strategy has become a core function for companies. A startup’s ability to understand where it fits into that strategy can determine whether a company expands, stalls, or dies.
Keith Figlioli is a managing partner at LRVHealth, a venture capital platform that invests in technology-based businesses in healthcare on behalf of health systems and payers.
