Why we build every SaaS platform serverless on AWS (and why it matters)
Serverless isn't just a cost optimization. It's an architectural choice that changes how you think about reliability, scaling, and operations. Here's how we think about it.
We've built payment platforms that process over $100M USD. We've built multi-tenant SaaS that serves 180,000+ users. Every one of them runs serverless on AWS.
This isn't a coincidence. It's a deliberate architectural choice we make at the start of every project.
Why serverless wins for SaaS
Traditional servers are provisioned for peak load. In practice, most SaaS platforms have highly variable traffic — bursts during business hours, near silence at night. With traditional infrastructure, you're paying for idle capacity.
Lambda, API Gateway, DynamoDB — these services scale to exactly what you need. Zero when you have zero traffic. Thousands of concurrent executions when you're under load.
The reliability argument
Serverless forces good architecture. If a Lambda function crashes, it doesn't bring down the whole system. Functions are isolated, retriable, and independently scalable.
For mission-critical systems — payment portals that governments rely on — this isolation is essential.
The operational reality
There's no server to patch. No capacity planning meeting. No 3am alert because a database server ran out of disk space.
The tradeoff is cold starts and the learning curve of thinking in functions. In 2026, cold starts are largely solved. The learning curve is real, but it's worth it.
Our stack
Lambda for compute, API Gateway for routing, DynamoDB for most storage, S3 for static assets, CloudFront for CDN. We layer Convex on top for real-time data needs.
It's not the most fashionable stack. But it works, reliably, at scale, every day.