Vertical SaaS Investor Luke Proposes Redrawing the Work Map in the AI Era
Luke Sophinos, founder and investor in the vertical SaaS field, pointed out in a post that traditional vertical software like Toast captures a complete work system through the restaurant operation map in its S-1, encompassing everything from customer flow, labor, to orders, inventory, payments, and management attention, rather than just functional modules. This was the core advantage of the previous generation of vertical SaaS.
He believes AI has changed this practice: old software records work outcomes, while the new generation of AI-native vertical software must first map actual "work" (messy labor, including interpretation, coordination, exception handling, and human judgment), rather than just recording clean "workflows." He summarized AI entry points using examples like Camber in medical reimbursement, OpenEvidence in clinical decision-making, and Datagrid's AI agents on top of systems, which include absorbing labor-intensive services, building industry-specific intelligent layers, or relying on existing record systems to build agents, thus helping vertical enterprises truly operate their businesses rather than just record them.
Source: Public Information
ABAB AI Insight
This analysis reveals a structural shift in vertical software from a recording era to an execution era. The lasting value of the Toast map lies in its view of restaurants as living operational systems rather than functional stacks. This systemic perspective has helped vertical SaaS companies establish data gravity and network effects in fragmented markets. However, the emergence of AI has broken technological boundaries: old software was limited by its inability to handle ambiguous inputs and real-time judgments, only able to store results after human judgment; AI can directly intervene at the labor layer, absorbing the coordination costs of managers acting as "human middleware," as well as implicit expenses like exception clearing and prioritization.
This shift corresponds to deep adjustments in productivity and distribution mechanisms. Over long historical cycles, technological substitution often first compresses repetitive labor and then reshapes power structures. The AI wedge strategy—starting from AI services to tap into the labor pool, forming industry-specific GPTs to create domain intelligent layers, and then building agents relying on existing systems—lowers the replacement costs for new entrants while allowing capital to price different types of labor more accurately. In vertical fields like restaurants, the interpretative and upgrade work previously handled by humans will gradually be outsourced to AI, leading to flatter management hierarchies and a concentration of residual labor on higher-value decision-making.
From an industrial migration perspective, this accelerates the evolution of software from general tools to vertical intelligent infrastructure. Early SaaS relied on standardized workflows to achieve scale, while the AI era depends on finely mapping the real consumption of "work" to establish defenses. Those who can grasp both high-level operational views and underlying labor details will gain lasting pricing power in data accumulation and decision-making loops, while most traditional vertical software faces the risk of being partially absorbed or restructured.
Overall, Luke Sophinos's observations point to a broader change in institutional and technological constraints: as software shifts from passive recording to active participation in judgment, the competitive focus in vertical markets shifts from functional coverage to labor absorption efficiency and system integration depth, reshaping the distribution of wealth among capital, founding teams, and executing labor.