While horizontal chatbots handle general tasks well, they fail at the highly specific, high-stakes workflows of professionals like investment bankers. Startups can build defensible businesses by creating opinionated products that master the final 1-2% of a use case, which provides significant value and is too niche for large AI labs to pursue.

The best application-focused AI companies are born from a need to solve a hard research problem to deliver a superior user experience. This "application-pull" approach, seen in companies like Harvey (RAG) and Runway (models), creates a stronger moat than pursuing research for its own sake.

Harvey's initial product was a tool for individual lawyers. The company found greater value by shifting focus to the productivity of entire legal teams and firms, tackling enterprise-level challenges like workflow orchestration, governance, and secure collaboration, which go far beyond simple model intelligence.

Successful vertical AI applications serve as a critical intermediary between powerful foundation models and specific industries like healthcare or legal. Their core value lies in being a "translation and transformation layer," adapting generic AI capabilities to solve nuanced, industry-specific problems for large enterprises.

While the "bitter lesson" suggests powerful general models will dominate, vertical AI solutions can thrive by deeply integrating with a company's specific data, workflows, and project context. The model can't know this proprietary information; value is created by the application that bridges this gap.

Initially building a tool for ML teams, they discovered the true pain point was creating AI-powered workflows for business users. This insight came from observing how first customers struggled with the infrastructure *around* their tool, not the tool itself.

Eve found Big Law wanted bespoke AI projects with marginal gains. In contrast, plaintiff firms had highly repeatable workflows where AI could drive massive efficiency, perfectly aligning with their contingency-fee business model, making them a far better target market.

The legal profession's core functions—researching case law, drafting contracts, and reviewing documents—are based on a large, structured corpus of text. This makes them ideal use cases for Large Language Models, fueling a massive wave of investment into legal AI companies.

Harvey intentionally avoids self-serve and focuses on the most complex enterprise legal work first. The strategy is to build a business around problems so difficult they will outlast the next decade of foundational model advancements, preventing commoditization.