AI is embedding itself into everyday tools and workflows, expanding who can build, decide, and operate at scale.

After years of skepticism about AI monetization, 2025 produced clearer proof points. AI-native startups like Cursor, Lovable, StackBlitz, and Emergent crossed $10 million in ARR within 18 months, driven by product-led adoption. At the same time, foundation model usage accelerated as research labs shipped increasingly capable models across modalities in rapid succession.

The market’s center of gravity has shifted from AI experiments to applied vertical AI, which has become the new battleground. Investors are targeting sector-specific use cases in healthcare, legal, code generation, customer support, and logistics. The difference lies in not just building wrappers around LLMs, but embedding AI into workflows to create product advantages.

Abridge illustrates the shift from novelty to operational impact in the healthcare sector. The platform transcribes and summarizes patient-provider conversations in real time, streamlining documentation into electronic health records and reducing the burden of after-hours clerical work, which is a major driver of clinician burnout. Staffing is frequently cited as the top concern for hospital leadership, and physicians spend substantial time on administrative work. Abridge’s value proposition is straightforward: reduce documentation time, improve accuracy, support real-time updates, and maintain compliance.

AI is beginning to operate as an active participant in organizational workflows, rather than just a passive interface. Instead of only handling simple, front-end interactions, AI systems are increasingly capable of understanding context, executing multi-step tasks, and coordinating across internal tools and teams. Decagon illustrates how, in customer operations, this shift changes the nature of work itself. The company’s customer support agents address each layer of the customer operations stack, resolving complex requests like refunds and shipment changes; continuously improving knowledge bases, and routing insights to product and engineering teams. Because poor customer experience carries real economic cost, the ability to handle nuanced workflows with reliability has become not just a support function, but a growth lever.

Vercel offers another example of how AI is reshaping everyday work rather than simply accelerating it. The company’s coding assistant, v0, reduces the friction between intent and output, allowing people to translate ideas into functional software with far fewer technical barriers. What matters is not just speed, but accessibility: capabilities that once required specialized expertise are increasingly available to a broader set of builders. This shift is broadening the scope of who can participate in software creation, and accelerating the rate at which products can move from concept to reality.

Together, these developments reflect the larger pattern we see across the economy: AI is no longer delivering incremental efficiency. Instead, it is quietly embedding itself into core tools and workflows, expanding workers’ capabilities, enabling new operating models and rebuilding industries.

As the cost of building software trends toward zero, value creation may shift partially back toward physical infrastructure and differentiated IP. At the same time, incumbents with durable data platforms are benefiting from renewed demand. Databricks, Snowflake, MongoDB, and Datadog have seen re-acceleration as enterprises look to trusted platforms to manage their data, orchestrate workflows, and monitor operations.

“Almost every single industry will be disrupted by AI. So far, AI has automated some of our most mundane work, but we look forward to the new creative ideas that can come out of AI…I think many of those experiences can be truly transformative in our lives.”

Alfred Lin,

Sequoia