Two startups at global scale without DevOps
By Steven Van ·
Leonardo.AI processes 4.5 million images a day and Relevance AI runs 50,000 agents, both without an in-house DevOps team.
A case study from Vercel profiles two Sydney-based AI startups, Leonardo.AI and Relevance AI, that run at global scale without a dedicated DevOps team.
- Leonardo.AI, an AI image generator for game developers, artists and marketers, now processes more than 4.5 million images a day. After moving to Vercel, its build times fell from over 10 minutes to 2, uncached page loads dropped from 60 seconds to about 3, and its product launch cycle shortened from months to two weeks. Co-founder and CTO Peter Runham said switching to Vercel "changed how we innovate."
- Relevance AI runs 50,000 agents that automate sales and marketing work across Salesforce, HubSpot and Slack, including agents that build, test and deploy web pages on their own using the Vercel REST API. In a v0 Power Hour session, Relevance AI's Scott Henderson built a similar agent in v0 and deployed it globally in a few clicks.
Vercel ties the pattern to a hiring gap across APAC: a DevOps engineer in Australia costs $150K or more and can take months to find, and IDC reports 60 to 80 percent of APAC organizations struggle to fill IT roles.
Vercel
The platform for frontend developers — deploy, preview, and scale web apps and AI agents with zero config.
View Vercel →