How Stably ships AI testing agents in hours, not weeks
By Steven Van ·
The 6-person startup runs its testing agents on Vercel's AI Gateway and Sandbox after winning Vercel's AI Accelerator.
Stably, a 6-person startup building AI agents that run autonomous end-to-end tests, says it cut the time to ship new testing agents from weeks to hours after moving its stack onto Vercel.
Stably's agents test applications by deploying on preview URLs, reading code diffs, and clicking through the app like a real user to check whether changes work. The company's frontend and backend run on Vercel, using AI Gateway for scaling AI requests and higher TPM limits, and Sandbox for agentic code execution.
According to the announcement, the move let Stably:
- Launch new product lines in hours instead of weeks
- Ship internal tools, such as Slack bots and dashboards, without separate deployment planning
- Scale from an early product to enterprise contracts on the same infrastructure
- Run its entire stack with a 6-person, revenue-generating team
Stably won Vercel's AI Accelerator along with an investment from Vercel Ventures, which it says helped build credibility in enterprise sales conversations. Since then, the company has started open-sourcing more of its core testing framework and internal tools.
