How OpenEvidence built a healthcare AI that physicians actually trust
Source: Vercel Blog
The Viral Moment
Andy Yoon, Lead Frontend Engineer, first noticed OpenEvidence’s surge when a TikTok video hit around two million views in less than a week. While the numbers climbed, the system remained stable: logs stayed green, response times were fast, and error rates stayed near zero.
“Vercel has just completely scaled with that usage,” Andy says. “We’ve never had it fall over due to capacity or had to provision anything extra. Just being able to trust that it’s there, to the point where we don’t really even think about it, is amazing.”
This proved that OpenEvidence had solved a problem many healthcare‑tech companies still wrestle with: moving at startup speed while meeting hospital‑grade reliability standards. A failure in a clinical decision‑support tool could lead to a bad medical decision, so stability, clarity, and trust are non‑negotiable.
Engineering Challenges
When Andy joined three years ago, he was essentially the only frontend engineer on a team dominated by Python and machine‑learning developers. The team needed infrastructure that required little babysitting—code could be deployed and would just work, scaling automatically as traffic grew.
Hybrid Architecture
- Backend: Python on Google Cloud Platform, handling data ingestion, model orchestration, and core business logic.
- Frontend: Next.js deployed on Vercel.
“Given the makeup of our engineering team, Vercel has really scaled with our frontend so well,” Andy notes.
Key workflow features:
- Every commit triggers an automatic deployment.
- Production deploys take about five minutes.
- Preview URLs are generated for every branch, enabling shareable live demos and instant rollbacks.
Rapid Prototyping and Product‑Market Fit
Early proof‑of‑concepts were each spun up as separate Vercel projects with custom domains. The process was simple:
- Spin up a new project.
- Connect a custom domain.
- Push code → a production‑like environment appears.
Stakeholders could instantly test workflows, accelerating the search for product‑market fit and helping secure early enterprise partnerships.
Scaling with Fluid Compute
As the platform grew 1,000×, lead infrastructure engineer Micah Smith monitored compute costs. Vercel’s Fluid Compute—which blends on‑demand execution with server‑like efficiency—was enabled, resulting in:
- 90 % reduction in serverless spend.
- Same reliability and performance.
- Faster response times and fewer cold starts.
“We reduced our serverless spend by 90 % while maintaining the same performance, and even as we’ve scaled up to 1,000× growth, Vercel is less than 5 % of our overall infra spend.” — Micah Smith, VP Engineering
Impact on Healthcare Professionals
Many clinicians still use outdated software that, while reliable, lacks modern usability. OpenEvidence had to “thread the needle” by delivering a contemporary, fast, and trustworthy solution that meets the high reliability bar expected in hospitals.
Since its viral moment, the platform now supports over 40 % of physicians in the United States, handling millions of clinical consultations daily across more than 10,000 hospitals and medical centers.
About OpenEvidence
OpenEvidence is the fastest‑growing clinical decision‑support platform in the United States and the most widely used medical search engine among U.S. clinicians. Trusted by hundreds of thousands of verified healthcare professionals, it provides point‑of‑care recommendations that are sourced, cited, and grounded in peer‑reviewed medical literature.
Learn more at .