The application layer is still early because most organizations have not redesigned their operations around AI. Adding a chatbot to an old process is only the beginning.
Near-term opportunities
The strongest near-term opportunities are repetitive, expensive workflows involving documents, calls, messages, decisions, and legacy systems. Buyers already understand the cost because they currently employ people or vendors to perform the work.
Beyond existing models
An application company can move deeper into the stack over time. It may collect unique data, train specialized models, create evaluation systems, develop new agent architectures, or build hardware for a particular environment.
From tool to institution
The largest application companies may not look like traditional software vendors. They may operate as AI-native accounting firms, clinics, logistics operators, research organizations, schools, or manufacturers.
The long-term opportunity is not merely to sell AI tools to an industry. It is to redesign how the industry produces value.
Questions for opportunity analysis
- Which work is costly, repetitive, and measurable?
- Where are customers dissatisfied with existing service?
- What data or learning loop improves performance over time?
- Which mistakes require human review?
- Can the company eventually move from assistance to responsibility for outcomes?