Developer-first browser automation to verticals
David Mlcoch, co-founder & CEO of Asteroid, on browser automation and the last mile problem of AI
This market is still being shaped by developer tools, which means the winning wedge is not raw browser control, but packaging reliability so an operations team can actually use it. Browserbase sits lower in the stack as hosted browser infrastructure plus Stagehand, an open source framework for writing automations, while Asteroid is pushing upward into a workflow product for insurance and healthcare teams that need repeated data entry into old web portals.
-
The pattern matches earlier automation markets. Browser automation started with QA tools like Selenium, then moved to Playwright as the standard developer framework. The current AI wave adds natural language control, but most products still assume someone technical can debug selectors, prompts, and browser sessions.
-
Browserbase and Browser Use both cater to builders first. Browserbase provides cloud browser sessions and positions Stagehand as the framework layer for AI web automation. Browser Use likewise began as a developer library, and later added hosted infrastructure so developers could run agents at scale.
-
Asteroid is targeting the budget that used to go to RPA projects and outsourced back office labor. That matters because UiPath scaled to $1.551B ARR by selling automation into enterprises, but often through technical implementation and services heavy deployments. Asteroid's bet is that AI can shrink that setup burden enough for domain experts to drive automation directly.
The next step in this market is a split between infrastructure and application layers. Infrastructure companies will supply browsers, sessions, and agent primitives, while product companies that encode vertical workflows, supervision, and repeatable deployment will capture the larger enterprise budget, especially in legacy heavy sectors like insurance and healthcare.