In a recent discussion, YC partners said the incubator was deluged with applications from startups building vertical AI agents, which could become the new SaaS
John Thornhill / Financial Times :
Context & Ripple Effects
YC had already identified AI as a major startup category: its winter 2024 cohort was reported as roughly half AI-focused, following an earlier vertical track for AI startups launched in 2017.
The application surge narrows that broad AI interest to agents built for particular industries and workflows. It is an early signal that founders see the application layer, rather than general-purpose models alone, as the place to build companies.
First-order effects
- YC must assess a much denser pool of companies targeting specialized AI workflows, while applicants face more direct competition from similarly positioned agent startups.
- The report gives vertical-agent founders a stronger incubator-market signal that their category is legible as a successor to conventional software products, even though it does not establish which companies will succeed.
Second-order effects
- As more startups target the same business processes, differentiation shifts toward workflow integration, proprietary customer context, and reliable execution rather than simply adding an AI interface.
- The concentration of agent builders can increase demand for underlying models and infrastructure, while making model providers more interchangeable where startups can switch among them—a possibility raised in coverage of agents weakening the strategic value of standalone LLMs.
Third-order effects
- If vertical agents consistently replace discrete software tasks, SaaS competition could shift from selling seats and features toward owning end-to-end operational workflows.
- YC's pipeline may help normalize an AI-native startup model in which specialized distribution and domain integration matter as much as access to a frontier model; later coverage of a heavily agentic YC demo-day batch suggests the category continued to expand.
The trend: Vertical AI agents are emerging as an application-layer push to turn general-purpose models into software that performs specific business work.