At nvp capital, we love thinking about opportunities in the “pen and paper” industries. We’ve always taken a more vertical approach to thesis building, but today there is an unprecedented opportunity to truly change how work happens through automation and workflow products solving problems in under-digitized verticals. This is driven by the convergence of improved data infrastructure, cheaper compute, and, of course, large language models.
We understand that true defensibility in many verticals comes not just from cutting-edge technology, but from a combination of a deep understanding of industry-specific nuances, strategic data utilization, and solving critical business challenges.
Drawing on insights from over 35 corporate partners, 15 vertical AI portfolio companies and countless conversations with founders, operators, and industry experts, we’ve identified key attributes that set high-potential start-ups apart. Here’s what we look for:
- Data: The Foundation of Sustainable Advantage
In vertical AI, data is more than an input—it’s the cornerstone of competitive differentiation. We target companies that leverage one or ideally all of the following data pillars:
Unstructured Data Mastery: We are excited about verticals that are rich in unstructured data – text, images, video, sensor readings, and more. Enterprises in under-digitized industries have spent years accumulating vast amounts of data, but only now—thanks to advancements in data infrastructure, LLMs, and lower computing costs—can they truly make sense of it. Roughly 80–90% of business data in enterprises is unstructured – and we believe that start-ups can earn the right to build workflow software and automate processes by effectively accessing, cleaning, organizing, and making sense of unstructured data. The more data a company ingests the more the model improves, providing acute and valuable insights.
Hard-to-Access Data as a Strategic Barrier: Data in sectors like law, finance and healthcare is often divided between internal records and public sources, with access further complicated by regulatory and legacy system constraints. Companies that can gain access to and integrate with these disparate sources begin to create a substantial barrier to entry. This access allows for the development of products and refinement of models that competitors cannot easily replicate. Of course, achieving this requires startups to first earn credibility and build trust with early customers. Establishing deep relationships with early adopters who are willing to share internal data in exchange for deep insights not only helps refine models but also contributes to helping the company build a truly unique and differentiated product.
Real-World Sensing: By capturing diverse real-world data streams (for example, camera feeds on factory floors, environmental sensors in agriculture, or wearable data in healthcare) and combining them with other modalities, companies can unlock richer insights and outcomes and create a compounding competitive advantage.
For example, our portfolio company TheoAI is building a legal prediction engine that leverages a multitude of data sources to quickly and accurately predict case outcomes. By mastering unstructured legal data, Theo transforms a process that once took hours or weeks into a streamlined, data-driven decision-making system, significantly reducing costs and increasing revenue.
- Building in Complex, Legacy Environments
Many high-impact vertical AI opportunities lie in complex, legacy industries—healthcare, finance, manufacturing, insurance, and beyond. These sectors are traditionally hard to break into due to complex compliance requirements, tough sales cycles, and complex integrations with legacy systems. However, these same challenges create compelling opportunities for startups that can navigate them effectively.
Delivering Clear ROI: The winners in these sectors deliver measurable returns—whether through driving revenue growth, reducing operational costs, or unlocking entirely new revenue streams.
Deep Industry Expertise: Especially in regulated industries, having a founder who truly understands the space is critical to overcoming regulatory hurdles and building trusted, compliant solutions.
Creating Ingrained Partnerships: Companies that solve mission-critical problems become indispensable partners. Their solutions, once implemented, are hard to replace not only because of the intrinsic value they provide but also due to the complexity of switching alternatives.
A great example of this is our company Upwell, which is automating invoicing and AR in the transportation industry. In order to deploy LLMs effectively they first have to integrate into a range of transportation management systems and invoice auditing ERPs, while building deep relationships with legacy players to earn their trust.
- Addressing Labor Shortages & Empowering the Worker
Many of these industries are grappling with significant labor shortages driven by an aging workforce, limited training capacity, and skill gaps exacerbated by rapid technological change. Critical roles are under threat: 23% of registered nurses plan to retire within five years, nearly 40% of attorneys are expected to retire in the next decade, and about 30% of skilled manufacturing workers are nearing retirement, straining operational capacity across sectors with new talent pipelines failing to keep pace. As the workforce shrinks, there are opportunities for startups to fill these labor gaps.
Empowering Knowledge Workers: AI can empower existing employees by taking on mundane work, streamlining research and analytics, and enabling strategic decision-making. We are especially compelled by solutions that make knowledge workers the superhero: enabling workers to achieve more with greater efficiency and higher impact, with AI taking on entire workflows of lower value, redundant tasks.
Driving Operational Efficiency and Revenue Growth: With fewer resources, organizations can leverage AI to scale their operations, ensuring critical functions remain robust even as talent pipelines falter. This enhanced efficiency not only improves day-to-day operations but also unlocks unprecedented revenue growth by enabling enterprises, large and small, to do more with less.
For example, in the realm of cybersecurity, Compyl—an end-to-end security and compliance platform leveraging GenAI and large language models—empowers IT specialists by automating continuous security and regulatory processes. This enables cybersecurity teams to shift their focus from routine tasks to strategic threat mitigation, directly addressing labor shortages in a high-demand field.
Similarly, in transportation and logistics, Class8 is redefining trucking by treating carriers as valued customers rather than just data points. Their solution streamlines operations and optimizes load management, effectively alleviating labor pressures and enhancing operational efficiency.
Both examples illustrate how targeted AI solutions not only compensate for labor shortages but also empower knowledge workers and drive significant growth across sectors.
If you’re a founder building a vertical AI company that aligns with these principles, we’d love to connect. We’re looking for businesses that combine deep industry expertise, proprietary data, and a relentless focus on solving mission-critical problems. If you are building in Vertical AI, please reach out to dan@nvpcap.com or skylar@nvpcap.com.
