• Agentic behaviour: AI products that autonomously initiate actions, moving beyond command-response architectures.
• Action orientation: Rather than a separate system type, "action" represents an orientation that can exist within any system (record, engagement, or work). It enables the execution of discrete tasks based on data and user inputs. In traditional software, we often referred to these as "actionable insights," which were often served up in dynamic dashboards, where the human user would then move offline to complete the "work." Modern systems increasingly incorporate actions directly, allowing users to complete tasks, close loops, and finish workflows without ever leaving the system where the information lives.
• AI moats: This is where the real defensibility in AI lies. Unlike traditional software that just piles up data, AI moats transform that data into something genuinely valuable. They add structure and context that generic LLMs simply can’t match. What’s exciting is how they capture the ‘grey’ knowledge → all that expertise and behavioural nuance that typically only exists in people’s heads. The best Vertical AI companies build moats through superior model capabilities and agentic workflows that get dramatically better with every interaction. We’re seeing these advantages compound over time, creating exponential value gaps that leave competitors scrambling to catch up, even when they have similar tech foundations. This isn’t just incremental improvement, it’s a fundamentally different approach to building lasting competitive advantage.
• Currency: The measurable financial impact a solution delivers to the customer, whether through margin expansion, cost reduction, or net new revenue creation. High-currency solutions expand budgets rather than compete for legacy spend. High-currency solutions have the capacity to actually increase the top-line revenue for a business customer, by producing more work, creating better products & services, or increasing market access.
• Customer’s customer: Solutions that create value downstream for the end-user or client of the buyer. Prioritising the customer’s customer increases strategic stickiness and positions the product closer to critical revenue flows and decision points.
• Data currency: The phenomenon where larger datasets attract applications, services and additional data toward them due to their mass. As data accumulates, it becomes more efficient to bring compute resources and applications to the data rather than moving data elsewhere. Healthcare EHR systems with vast patient histories naturally attract clinical applications, while legacy systems create migration challenges because of their significant gravitational pull. This effect influences architectural decisions and creates natural centralisation that can evolve into critical workflow hubs.
• Focal point: The critical junction in a customer’s workflow that is both mission-critical and offers the greatest expansion potential. These are the "unswitchoffable" systems that, once embedded, provide significant leverage for retention and pricing power. In traditional software, Systems of Record often became control points because they were painful to replace, but in the AI era, focal points increasingly form around workflows that directly impact key business metrics. By owning this strategic position in the customer’s operation, companies can expand into adjacent areas and command premium pricing.
• Front & back office convergence: Traditional software often sat in the back office, where the desk workers reigned. Multi-modal AI (voice, audio, image) can collapse traditional operational silos by embedding actionable insights and task execution into frontline workflows. Front-to-back integration enables workers to execute tasks without functional boundaries.
• Land and expand motion: Beginning with a narrowly scoped, high-value wedge into a workflow and expanding outward through adjacent capability development. This motion leverages initial user trust to systematically erode legacy systems and increase control over vertical-specific workflows. This was a common strategy used in traditional software and is even more useful for AI-native market entrants.
• Model/market fit (MMF): Achieving scalable monetisation strategies that align with the value created in critical workflows. In Vertical AI, MMF often requires moving beyond traditional SaaS seat models to pricing based on workflow ownership, value capture, or direct productivity gains.
• OTT (over-the-top) strategy: Building value by integrating across existing incumbent systems without requiring displacement, reducing adoption friction. OTT layers evolve into default operational interfaces and, over time, can collapse incumbent system relevance without overt replacement. A transcription tool, that surfaces proprietary data and works nicely with the existing control point, is an example of an OTT strategy to enter a market. This is a common market entry strategy for AI-native tools today, as they build trust with the user to offer a more expanded ‘work-producing’ proposition.
• Product/market fit (PMF): Building products that solve critical, painful workflow problems so effectively that users demand adoption and usage spreads organically. In Vertical AI, PMF often shows through high retention, deep "embeddedness" into revenue-critical processes, and organic workflow expansion.
• Proprietary data moats: Unique data that is captured through interaction with the platform. The generation of high-quality data creates barriers to replication that widen over time with usage and integration into the customer’s systems. Sometimes this data exists, but it’s offline or unstructured, with no context. Sometimes it doesn’t exist except inside the user’s head, and needs to be captured.
• RTBD (role to be done): Extending beyond "jobs to be done" to AI systems that assume full professional roles across workflows. RTBD unlocks full workflow ownership opportunities, enabling pricing and defensibility levels unavailable to task-specific AI or traditional SaaS models.
• System of engagement: User-facing tools designed for active interaction. We’re already seeing these evolving from static interfaces into personalised experiences that drive actions, data generation, and loyalty loops. We often see this in AI-native companies as a chat interface. They serve as the bridge between users and the underlying systems, representing the communication layer that connects people with data and processes, similar to how co-pilots augment human capabilities rather than replacing them outright.
• System of record: These are authoritative databases that store structured, operationally critical information. While they’ve traditionally held defensible value, they risk being commoditised as AI-native systems of work and action increasingly take on the role of ’source of truth’—and use that data to drive higher-value outcomes.

• System of work: Platforms that actively perform core productive tasks rather than simply storing information (Salesforce) or enabling communication (Slack). Systems of work have the potential to replace human inputs over time, which would fundamentally shift the economics of operating a business. A software providing AI agents that produce actual work (e.g., a marketing document, a report, or an entire software application), autonomously, is a System of Work. Within a System of Work, there may be multiple other systems working in harmony together to produce the work. These systems represent a genuine threat to all incumbent systems by focusing on autonomous work production. Importantly, these systems aren’t strictly evolutionary; they co-exist with other system types depending on category or sector. However, Systems of Work challenge everything before them by fundamentally changing how work gets done, and they’ll continue to transform as AI capabilities advance.
• Task-based vs software-based pricing: Transitioning from static licence models to outcome-aligned monetisation structures. In Vertical AI, task-completion pricing models unlock budget pools tied to operational expenditure rather than IT procurement, enabling faster sales cycles and broader market access. As an example, AI products built to resolve customer queries may be priced based on the number of resolved queries.
• Technology in underserved markets: Industries where traditional software failed to provide adoption-worthy value, leaving workflows reliant on manual processes or legacy solutions. Vertical AI can leapfrog these markets by delivering immediate, tangible operational improvements.
• Vertical-specific domain expertise: Deep practitioner-level knowledge of an industry’s workflows, compliance structures, and operational edge cases. Founders with embedded domain insight can build products that achieve faster PMF, lower friction adoption, and higher defensibility against horizontal entrants.
• White space opportunities: Entirely new workflow layers or categories unlocked by AI capabilities, not previously accessible through manual or traditional software approaches. White space strategies define new standards and enjoy a first-mover advantage in industry process redefinition.
• Worker experience (WX): Systems that radically improve frontline worker output and job quality, serving as competitive moats in adoption, retention, and advocacy. In Vertical AI, and in particular with Systems of Work, WX emerges as a core driver of bottom-up adoption motions across critical operational verticals.
• Workflow criticality: How central a workflow is to a company’s success. The more essential it is, the harder it is to change, but the more valuable it is to improve. Workflows like patient billing or asset inspections exemplify mission-critical workflows. If an AI-native product can plug in and add value to workflows, essential meeting business KPIs, they become harder to replace.

Closing reflection

The AI landscape is evolving rapidly, reshaping not just technology but the very foundations of work itself. At Tidal Ventures, we believe the most successful Vertical AI companies will not simply automate tasks, but embed intelligence deep into workflows, unlock new economic layers, and redefine how industries operate.

The concepts captured in this glossary are our way of describing the building blocks of the next generation of global technology leaders. As the frontier shifts, we will continue expanding and refining it. We also know that these concepts sometimes come across as really abstract, so we’re going to do some work to provide more actionable examples as this evolves.

If you are building at the intersection of work, systems, and AI, we would love to hear from you.